frc-scheduler-server: pr-fix-known-vulns-run5

What happened, why, and what came out of it. Written by the harness from its own records; a person decides what to do with it. The same story opens the run's folder as README.md; the example's write-up is on its example page.

repository frc-scheduler-serverlanguage pythonmode diffbase 2ad04243861bhead cfa8f4a3292bmodel qwen/qwen3.8-27bstarted 2026-09-12T01:41:41+00:00harness 0.1.0run 76a3fa863e76

pyasn1. This change moves pyasn1 from 0.6.4 to 0.4.8, a version with 4 known vulnerabilities. That is a downgrade. The harness proved 1 of the 4 with a test that fails on the vulnerable version and passes on the fixed one; the other 3 are unproven and marked so. Accept the 10 candidate tests if they look right, act on the findings below, and send the draft VEX statements to Product Security.

python-jose. This change updates python-jose from 3.3.0 to 3.4.0, which closes 3 known vulnerabilities. The harness proved 1 of the 3 with a test that fails on the vulnerable version and passes on the fixed one; the other 2 are unproven and marked so. Accept the 7 candidate tests if they look right, act on the findings below, and send the draft VEX statements to Product Security.

starlette. This change leaves starlette at 0.41.3, which has 7 known vulnerabilities the application is exposed to. The harness proved 1 of the 7 with a test that fails on the vulnerable version and passes on the fixed one; the other 6 are unproven and marked so. Accept the 6 candidate tests if they look right, act on the findings below, and send the draft VEX statements to Product Security.

Who, what, why, where, when

Who	The AI Test Harness 0.1.0 did the work without a person in the loop. A person decides the result; nothing in this run has been merged.
What	Tests for pyasn1 0.6.4 to 0.4.8, python-jose 3.3.0 to 3.4.0, starlette 0.41.3, chosen from a work list of 64 packages, 5 of which this change touched. Repository frc-scheduler-server, change 2ad04243861b..cfa8f4a3292b, mode diff.
Why	The risk score decides the budget; the two strongest reasons per package: pyasn1 (score 90, budget full): reachability unknown, treated as reachable; 8 known vulnerabilities for this version; 0 advisories before the change, 8 after. python-jose (score 51, budget full): reachable from first-party code; 159 source lines changed; 5 advisories before the change, 0 after. starlette (score 80, budget full): reachable from first-party code; 14 known vulnerabilities for this version.
Where	Every test ran in a sealed sandbox (podman, network none, image docker.io/library/python:3.12-slim); the model was qwen/qwen3.8-27b in agent mode; the run's own files are in this folder.
When	Started 2026-09-12 01:41:41 UTC. The last stage to write was assessment at 2026-10-13 18:34:45 UTC.

Decisions the harness made

pyasn1: cut 2 generated test(s) before execution: 2 fails on baseline after repair.
pyasn1: triage decided per test: 2 accept as CVE evidence; VEX status affected: the change introduces the exposure; 4 discard as CVE evidence; keep as characterization only if reviewer wants it; 2 open issue; reproducer is the test; advisory stays under investigation; 8 accept as characterization candidate.
pyasn1: finding, security at confidence 0.95: pyasn1 downgraded 0.6.4 -> 0.4.8 into a version with 8 known advisories. Route: Product Security.
pyasn1: finding, security at confidence 0.75: pyasn1 0.4.8 has 8 open advisories at head; reachable=unknown. Route: Product Security.
pyasn1: finding, defect at confidence 0.8: pyasn1: a code path raised the same internal error on both versions during CVE test generation (CVE-2026-59885). Route: open issue with reproducer.
pyasn1: finding, redundant at confidence 0.5: pyasn1: test_printer_is_callable killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
pyasn1: finding, redundant at confidence 0.5: pyasn1: test_debug_is_callable killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
pyasn1: finding, redundant at confidence 0.5: pyasn1: test_setlogger_returns_none killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
pyasn1: finding, redundant at confidence 0.5: pyasn1: test_hexdump_empty_bytes killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
pyasn1: finding, redundant at confidence 0.5: pyasn1: test_pyasn1error_is_exception_subclass killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
pyasn1: finding, redundant at confidence 0.5: pyasn1: test_value_constraint_error_is_pyasn1error_subclass killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
pyasn1: finding, redundant at confidence 0.5: pyasn1: test_substrate_underrun_error_is_pyasn1error_subclass killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
pyasn1: finding, redundant at confidence 0.5: pyasn1: test_pyasn1unicode_error_is_unicode_error_subclass killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
pyasn1: mutation score 0.417 is below the blueprint's 0.6; the packet says so, and the reviewer sees which tests killed nothing.
pyasn1: draft VEX per advisory: CVE-2026-30922 affected, CVE-2026-59884 under_investigation, CVE-2026-59885 under_investigation, CVE-2026-59886 under_investigation.
python-jose: cut 5 generated test(s) before execution: 5 fails on baseline after repair.
python-jose: triage decided per test: 2 open issue; reproducer is the test; advisory stays under investigation; 2 discard as CVE evidence; keep as characterization only if reviewer wants it; 2 accept as CVE evidence; VEX status fixed; 5 accept as characterization candidate.
python-jose: finding, defect at confidence 0.4: python-jose: a code path raised the same internal error on both versions during CVE test generation (CVE-2024-33663); NOT corroborated by stage 4 verdicts. Route: escalate: below confidence threshold.
python-jose: finding, defect at confidence 0.8: python-jose: a code path raised the same internal error on both versions during CVE test generation (CVE-2024-29370). Route: open issue with reproducer.
python-jose: finding, redundant at confidence 0.6: python-jose: test_jwt_decode_wrong_key_raises_jwterror's 2 mutant kill(s) are all among test_jwt_decode_expired_raises_expiredsignatureerror's 4. Route: escalate: below confidence threshold.
python-jose: finding, redundant at confidence 0.6: python-jose: test_jwt_get_unverified_header's 2 mutant kill(s) are all among test_jwt_decode_expired_raises_expiredsignatureerror's 4. Route: escalate: below confidence threshold.
python-jose: finding, redundant at confidence 0.6: python-jose: test_jwt_get_unverified_claims's 2 mutant kill(s) are all among test_jwt_decode_expired_raises_expiredsignatureerror's 4. Route: escalate: below confidence threshold.
python-jose: finding, redundant at confidence 0.6: python-jose: test_jwt_decode_expired_raises_expiredsignatureerror's 4 mutant kill(s) are all among test_jwt_encode_decode_roundtrip's 5. Route: escalate: below confidence threshold.
python-jose: mutation score 0.417 is below the blueprint's 0.6; the packet says so, and the reviewer sees which tests killed nothing.
python-jose: draft VEX per advisory: CVE-2024-29370 under_investigation, CVE-2024-33663 under_investigation, CVE-2024-33664 fixed.
starlette: cut 6 generated test(s) before execution: 6 fails on baseline after repair.
starlette: triage decided per test: 4 regenerate or discard; 6 discard as CVE evidence; keep as characterization only if reviewer wants it; 2 open issue; reproducer is the test; advisory stays under investigation; 2 accept as CVE evidence; VEX status affected: the change introduces the exposure; 4 accept as characterization candidate.
starlette: finding, security at confidence 0.9: starlette 0.41.3 has 14 open advisories at head; reachable=true. Route: Product Security.
starlette: finding, defect at confidence 0.4: starlette: a code path raised the same internal error on both versions during CVE test generation (CVE-2025-54121); NOT corroborated by stage 4 verdicts. Route: escalate: below confidence threshold.
starlette: finding, defect at confidence 0.4: starlette: a code path raised the same internal error on both versions during CVE test generation (CVE-2026-54283); NOT corroborated by stage 4 verdicts. Route: escalate: below confidence threshold.
starlette: finding, defect at confidence 0.8: starlette: a code path raised the same internal error on both versions during CVE test generation (CVE-2026-48817). Route: open issue with reproducer.
starlette: finding, obsolete at confidence 0.85: starlette: test_ghsa_jp82_jpqv_5vv3_exposure references starlette.requests.Request, which the upgrade to 1.3.1 changes the required parameters of. Route: propose retirement.
starlette: finding, redundant at confidence 0.5: starlette: test_simple_user_authenticated killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
starlette: finding, redundant at confidence 0.5: starlette: test_cookie_parser_basic killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
starlette: finding, redundant at confidence 0.5: starlette: test_response_status_and_body killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
starlette: finding, redundant at confidence 0.5: starlette: test_json_response_renders_json killed none of 12 sampled mutants in this run. Route: escalate: below confidence threshold.
starlette: mutation score 0.0 is below the blueprint's 0.6; the packet says so, and the reviewer sees which tests killed nothing.
starlette: draft VEX per advisory: CVE-2025-54121 affected, CVE-2025-62727 affected, CVE-2026-48710 affected, CVE-2026-48817 affected, CVE-2026-48818 affected, CVE-2026-54282 affected, CVE-2026-54283 affected.

Actions it took

pyasn1: wrote a review packet with 10 accepted test(s) as a patch, ready for a pull request.
pyasn1: drafted VEX statements for Product Security.
pyasn1: signed the test-result statement (PASSED).
pyasn1: no pull request was opened in this run; the packet holds what one would carry.
python-jose: wrote a review packet with 7 accepted test(s) as a patch, ready for a pull request.
python-jose: drafted VEX statements for Product Security.
python-jose: signed the test-result statement (PASSED).
python-jose: no pull request was opened in this run; the packet holds what one would carry.
starlette: wrote a review packet with 6 accepted test(s) as a patch, ready for a pull request.
starlette: drafted VEX statements for Product Security.
starlette: signed the test-result statement (PASSED).
starlette: no pull request was opened in this run; the packet holds what one would carry.

Actions it did not take, by design

It did not merge anything. Tests reach a repository only through a pull request that a person merges.
It did not change production code and did not propose a fix. Fixes come from a separate fix pipeline; this harness only tests.
It did not publish a VEX statement. The drafts are proposals for Product Security.
It did not delete any existing test. Retirements are proposed with evidence for a person to act on.

How it happened, stage by stage

What each stage read, did and wrote, with the file behind every number.

Self-verificationstage 0
Read
the harness's own code and the failure register
Did
- 28 register checks, all passed; sandbox probe skipped; model probe skipped.
Wrote
selfcheck/selfcheck.json
Intakestage 1
Read
the repository at base and head, its dependency manifest
Did
- Resolved 63 packages at head and 63 at base with ; looked every version up in OSV: 4 with advisories at head; 5 changed; work list of 64 rows.
Wrote
intake/worklist.json · intake/graph.new.json · intake/sbom.new.cdx.json · intake/vulns.json
Analysisstage 2
Read
the work list; package source, advisories, the application's call sites
Did
- 3 package(s) analyzed: pyasn1 score 90, budget full, 8 advisories, 146 breaking changes; python-jose score 51, budget full, 0 advisories, 0 breaking changes; starlette score 80, budget full, 14 advisories, 0 breaking changes.
Wrote
analyze/pyasn1/facts.json · analyze/python-jose/facts.json · analyze/starlette/facts.json
Generationstage 3
Read
the fact bundle per package; the model; the sealed sandbox for baseline runs
Did
- Model qwen/qwen3.8-27b, mode agent.
- pyasn1: 16 tests kept in 5 file(s), 47 model calls.
- python-jose: 11 tests kept in 4 file(s), 34 model calls.
- starlette: 18 tests kept in 8 file(s), 85 model calls.
Wrote
generate/pyasn1/manifest.json · tests/ · model-calls/ · generate/python-jose/manifest.json · tests/ · model-calls/ · generate/starlette/manifest.json · tests/ · model-calls/
Validation executionstage 4
Read
the candidate tests; the environments at head, base or the fixed candidate
Did
- pyasn1: sandbox runs new, new-rerun, old on podman (image digest not recorded); 16 tests, 12 pass on head, 0 flaky, 2 fix-pinning confirmed; 1423 lines of the package covered.
- Mutation: 385 sites on executed lines, 12 sampled, 5 killed, score 0.417.
- Relevance: 16 generated and 47 application tests examined; 0 obsolete, 8 redundant proposed, horizon this change.
- python-jose: sandbox runs new, new-rerun, old on podman (image digest not recorded); 11 tests, 9 pass on head, 0 flaky, 2 fix-pinning confirmed; 640 lines of the package covered.
- Mutation: 157 sites on executed lines, 12 sampled, 5 killed, score 0.417.
- Relevance: 11 generated and 47 application tests examined; 0 obsolete, 4 redundant proposed, horizon this change.
- starlette: sandbox runs new, new-rerun, old on podman (image digest not recorded); 18 tests, 10 pass on head, 0 flaky, 2 fix-pinning confirmed; 1368 lines of the package covered.
- Mutation: 319 sites on executed lines, 12 sampled, 0 killed, score 0.0.
- Relevance: 18 generated and 47 application tests examined; 1 obsolete, 4 redundant proposed, horizon the upgrade to 1.3.1.
Wrote
execute/pyasn1/results.json · new/ · execute/pyasn1/mutation/mutation.json · execute/pyasn1/relevance.json · execute/python-jose/results.json · new/ · execute/python-jose/mutation/mutation.json · execute/python-jose/relevance.json · execute/starlette/results.json · new/ · execute/starlette/mutation/mutation.json · execute/starlette/relevance.json
Triagestage 5
Read
every verdict and finding from stage 4
Did
- pyasn1: 11 findings, 10 tests to accept, 6 to discard or regenerate, 8 escalated to a person (threshold 0.7).
- python-jose: 6 findings, 7 tests to accept, 4 to discard or regenerate, 5 escalated to a person (threshold 0.7).
- starlette: 9 findings, 6 tests to accept, 12 to discard or regenerate, 6 escalated to a person (threshold 0.7).
Wrote
triage/pyasn1/triage.json · triage/python-jose/triage.json · triage/starlette/triage.json
Packetstage 6
Read
triage, the tests, the advisories
Did
- pyasn1: 10 accepted tests in the patch; VEX drafts {"CVE-2026-30922": "affected", "CVE-2026-59884": "under_investigation", "CVE-2026-59885": "under_investigation", "CVE-2026-59886": "under_investigation"}.
- python-jose: 7 accepted tests in the patch; VEX drafts {"CVE-2024-29370": "under_investigation", "CVE-2024-33663": "under_investigation", "CVE-2024-33664": "fixed"}.
- starlette: 6 accepted tests in the patch; VEX drafts {"CVE-2025-54121": "affected", "CVE-2025-62727": "affected", "CVE-2026-48710": "affected", "CVE-2026-48817": "affected", "CVE-2026-48818": "affected", "CVE-2026-54282": "affected", "CVE-2026-54283": "affected"}.
Wrote
packet/pyasn1/packet.md · packet/pyasn1/tests.patch · packet/pyasn1/vex.openvex.json · packet/python-jose/packet.md · packet/python-jose/tests.patch · packet/python-jose/vex.openvex.json · packet/starlette/packet.md · packet/starlette/tests.patch · packet/starlette/vex.openvex.json
Attestationstage attest
Read
the manifest, the tests, the results; UDLM's chain code
Did
- pyasn1: 10 provenance records; statement PASSED, key sha256:4f2defc2e375, verified at build; UDLM records sealed, schema problems 0.
- python-jose: 7 provenance records; statement PASSED, key sha256:4f2defc2e375, verified at build; UDLM records sealed, schema problems 0.
- starlette: 6 provenance records; statement PASSED, key sha256:4f2defc2e375, verified at build; UDLM records sealed, schema problems 0.
Wrote
attest/pyasn1/statement.dsse.json · attest/pyasn1/MANIFEST.json · udlm/ · attest/python-jose/statement.dsse.json · attest/python-jose/MANIFEST.json · udlm/ · attest/starlette/statement.dsse.json · attest/starlette/MANIFEST.json · udlm/
Assessmentstage assess
Read
every record the run wrote, against the blueprint's goals
Did
- 10 goals met, 1 not met, 1 not applicable, 1 measured without a target.
Wrote
assess/assess.md
Test pull requeststage propose
Read
the packet and the records
Did
- did not run in this run.
Wrote
nothing
Feedbackstage 7
Read
the pull request as a person decided it
Did
- did not run in this run.
Wrote
nothing

pyasn1 0.6.4 → 0.4.8

In plain terms. This change moves pyasn1 from 0.6.4 to 0.4.8, a version with 4 known vulnerabilities. That is a downgrade. The harness proved 1 of the 4 with a test that fails on the vulnerable version and passes on the fixed one; the other 3 are unproven and marked so. Accept the 10 candidate tests if they look right, act on the findings below, and send the draft VEX statements to Product Security.

Tests, one row each

Test	Category	Old	New	Verdict	Action
`test_ghsa_jr27_m4p2_rc6r_fix_pinning`	cve	pass	fail	exposure confirmed: the downgrade to the vulnerable version fails this test and the previous, fixed version passes it	accept as CVE evidence; VEX status affected: the change introduces the exposure
`test_ghsa_jr27_m4p2_rc6r_exposure`	cve	pass	fail	exposure confirmed: the downgrade to the vulnerable version fails this test and the previous, fixed version passes it	accept as CVE evidence; VEX status affected: the change introduces the exposure
`test_ghsa_m4p7_r5rc_7g4j_fix_pinning`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_m4p7_r5rc_7g4j_exposure`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_8ppf_4f7h_5ppj_fix_pinning`	cve	fail	fail	blocked on both versions: ValueError on old and new; possible package or environment defect, reproducer attached	open issue; reproducer is the test; advisory stays under investigation
`test_ghsa_8ppf_4f7h_5ppj_exposure`	cve	fail	fail	blocked on both versions: ValueError on old and new; possible package or environment defect, reproducer attached	open issue; reproducer is the test; advisory stays under investigation
`test_ghsa_hm4w_wwcw_mr6r_fix_pinning`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_hm4w_wwcw_mr6r_exposure`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_printer_is_callable`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_debug_is_callable`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_setlogger_returns_none`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_hexdump_empty_bytes`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_pyasn1error_is_exception_subclass`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_value_constraint_error_is_pyasn1error_subclass`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_substrate_underrun_error_is_pyasn1error_subclass`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_pyasn1unicode_error_is_unicode_error_subclass`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate

Findings

Class	Confidence	Finding	Route
security	0.95	pyasn1 downgraded 0.6.4 -> 0.4.8 into a version with 8 known advisories	Product Security
security	0.75	pyasn1 0.4.8 has 8 open advisories at head; reachable=unknown	Product Security
defect	0.8	pyasn1: a code path raised the same internal error on both versions during CVE test generation (CVE-2026-59885)	open issue with reproducer
redundant	0.5	pyasn1: test_printer_is_callable killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	pyasn1: test_debug_is_callable killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	pyasn1: test_setlogger_returns_none killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	pyasn1: test_hexdump_empty_bytes killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	pyasn1: test_pyasn1error_is_exception_subclass killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	pyasn1: test_value_constraint_error_is_pyasn1error_subclass killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	pyasn1: test_substrate_underrun_error_is_pyasn1error_subclass killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	pyasn1: test_pyasn1unicode_error_is_unicode_error_subclass killed none of 12 sampled mutants in this run	escalate: below confidence threshold

Draft VEX statements, for Product Security

Vulnerability	Status	Basis
CVE-2026-59885	under_investigation	open at head; reachable=unknown; fixed in ['0.6.4']
CVE-2026-59886	under_investigation	open at head; reachable=unknown; fixed in ['0.6.4']
CVE-2026-30922	affected	exposure test(s) ['test_ghsa_jr27_m4p2_rc6r_fix_pinning', 'test_ghsa_jr27_m4p2_rc6r_exposure'] fail on pkg:pypi/pyasn1@0.4.8 and pass on pkg:pypi/pyasn1@0.6.4: the change introduce
CVE-2026-59884	under_investigation	open at head; reachable=unknown; fixed in ['0.6.4']

Mutation: which test killed what

385 sites, 12 sampled, 5 killed, score 0.417.

Test	Mutants killed	Killed only by this test
`test_ghsa_hm4w_wwcw_mr6r_exposure`	3	0
`test_ghsa_hm4w_wwcw_mr6r_fix_pinning`	3	0
`test_ghsa_m4p7_r5rc_7g4j_exposure`	1	0
`test_ghsa_m4p7_r5rc_7g4j_fix_pinning`	1	0
`tests.test_pyasn1_cve_cve_2026_59884`	2	0
`tests.test_pyasn1_cve_cve_2026_59886`	2	0

Retirement proposals

Test	Reason	Evidence
`test_printer_is_callable`	redundant-no-kills	test_printer_is_callable killed none of 12 sampled mutants in this run
`test_debug_is_callable`	redundant-no-kills	test_debug_is_callable killed none of 12 sampled mutants in this run
`test_setlogger_returns_none`	redundant-no-kills	test_setlogger_returns_none killed none of 12 sampled mutants in this run
`test_hexdump_empty_bytes`	redundant-no-kills	test_hexdump_empty_bytes killed none of 12 sampled mutants in this run
`test_pyasn1error_is_exception_subclass`	redundant-no-kills	test_pyasn1error_is_exception_subclass killed none of 12 sampled mutants in this run
`test_value_constraint_error_is_pyasn1error_subclass`	redundant-no-kills	test_value_constraint_error_is_pyasn1error_subclass killed none of 12 sampled mutants in this run
`test_substrate_underrun_error_is_pyasn1error_subclass`	redundant-no-kills	test_substrate_underrun_error_is_pyasn1error_subclass killed none of 12 sampled mutants in this run
`test_pyasn1unicode_error_is_unicode_error_subclass`	redundant-no-kills	test_pyasn1unicode_error_is_unicode_error_subclass killed none of 12 sampled mutants in this run

python-jose 3.3.0 → 3.4.0

In plain terms. This change updates python-jose from 3.3.0 to 3.4.0, which closes 3 known vulnerabilities. The harness proved 1 of the 3 with a test that fails on the vulnerable version and passes on the fixed one; the other 2 are unproven and marked so. Accept the 7 candidate tests if they look right, act on the findings below, and send the draft VEX statements to Product Security.

Tests, one row each

Test	Category	Old	New	Verdict	Action
`test_pysec_2025_185_fix_pinning`	cve	fail	fail	blocked on both versions: jose.exceptions.JWEError on old and new; possible package or environment defect, reproducer attached	open issue; reproducer is the test; advisory stays under investigation
`test_pysec_2025_185_exposure`	cve	fail	fail	blocked on both versions: jose.exceptions.JWEError on old and new; possible package or environment defect, reproducer attached	open issue; reproducer is the test; advisory stays under investigation
`test_ghsa_6c5p_j8vq_pqhj_fix_pinning`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_6c5p_j8vq_pqhj_exposure`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_cjwg_qfpm_7377_fix_pinning`	cve	fail	pass	fix-pinning confirmed: fails on vulnerable, passes on fixed	accept as CVE evidence; VEX status fixed
`test_ghsa_cjwg_qfpm_7377_exposure`	cve	fail	pass	fix-pinning confirmed: fails on vulnerable, passes on fixed	accept as CVE evidence; VEX status fixed
`test_jwt_encode_decode_roundtrip`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_jwt_decode_wrong_key_raises_jwterror`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_jwt_get_unverified_header`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_jwt_get_unverified_claims`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_jwt_decode_expired_raises_expiredsignatureerror`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate

Findings

Class	Confidence	Finding	Route
defect	0.4	python-jose: a code path raised the same internal error on both versions during CVE test generation (CVE-2024-33663); NOT corroborated by stage 4 verdicts	escalate: below confidence threshold
defect	0.8	python-jose: a code path raised the same internal error on both versions during CVE test generation (CVE-2024-29370)	open issue with reproducer
redundant	0.6	python-jose: test_jwt_decode_wrong_key_raises_jwterror's 2 mutant kill(s) are all among test_jwt_decode_expired_raises_expiredsignatureerror's 4	escalate: below confidence threshold
redundant	0.6	python-jose: test_jwt_get_unverified_header's 2 mutant kill(s) are all among test_jwt_decode_expired_raises_expiredsignatureerror's 4	escalate: below confidence threshold
redundant	0.6	python-jose: test_jwt_get_unverified_claims's 2 mutant kill(s) are all among test_jwt_decode_expired_raises_expiredsignatureerror's 4	escalate: below confidence threshold
redundant	0.6	python-jose: test_jwt_decode_expired_raises_expiredsignatureerror's 4 mutant kill(s) are all among test_jwt_encode_decode_roundtrip's 5	escalate: below confidence threshold

Draft VEX statements, for Product Security

Vulnerability	Status	Basis
CVE-2024-33663, CVE-2026-85394	under_investigation	bump to a fixed version per advisory metadata; no confirmed fix-pinning test yet
CVE-2024-33664	fixed	fix-pinning test(s) ['test_ghsa_cjwg_qfpm_7377_exposure', 'test_ghsa_cjwg_qfpm_7377_fix_pinning'] fail on pkg:pypi/python-jose@3.3.0 and pass on pkg:pypi/python-jose@3.4.0
CVE-2024-29370	under_investigation	bump to a fixed version per advisory metadata; no confirmed fix-pinning test yet

Mutation: which test killed what

157 sites, 12 sampled, 5 killed, score 0.417.

Test	Mutants killed	Killed only by this test
`test_ghsa_cjwg_qfpm_7377_fix_pinning`	1	0
`test_jwt_decode_expired_raises_expiredsignatureerror`	4	0
`test_jwt_decode_wrong_key_raises_jwterror`	2	0
`test_jwt_encode_decode_roundtrip`	5	1
`test_jwt_get_unverified_claims`	2	0
`test_jwt_get_unverified_header`	2	0

Retirement proposals

Test	Reason	Evidence
`test_jwt_decode_wrong_key_raises_jwterror`	redundant-subsumed	test_jwt_decode_wrong_key_raises_jwterror's 2 mutant kill(s) are all among test_jwt_decode_expired_raises_expiredsignatureerror's 4
`test_jwt_get_unverified_header`	redundant-subsumed	test_jwt_get_unverified_header's 2 mutant kill(s) are all among test_jwt_decode_expired_raises_expiredsignatureerror's 4
`test_jwt_get_unverified_claims`	redundant-subsumed	test_jwt_get_unverified_claims's 2 mutant kill(s) are all among test_jwt_decode_expired_raises_expiredsignatureerror's 4
`test_jwt_decode_expired_raises_expiredsignatureerror`	redundant-subsumed	test_jwt_decode_expired_raises_expiredsignatureerror's 4 mutant kill(s) are all among test_jwt_encode_decode_roundtrip's 5

starlette 0.41.3 0.41.3

In plain terms. This change leaves starlette at 0.41.3, which has 7 known vulnerabilities the application is exposed to. The harness proved 1 of the 7 with a test that fails on the vulnerable version and passes on the fixed one; the other 6 are unproven and marked so. Accept the 6 candidate tests if they look right, act on the findings below, and send the draft VEX statements to Product Security.

Tests, one row each

Test	Category	Old	New	Verdict	Action
`test_GHSA_2c2j_9gv5_cj73_fix_pinning`	cve	fail	fail	fails on the fixed version: test bug or the advisory is misread; back to generation	regenerate or discard
`test_GHSA_2c2j_9gv5_cj73_exposure`	cve	fail	fail	fails on the fixed version: test bug or the advisory is misread; back to generation	regenerate or discard
`test_ghsa_7f5h_v6xp_fcq8_fix_pinning`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_7f5h_v6xp_fcq8_exposure`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_86qp_5c8j_p5mr_fix_pinning`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_86qp_5c8j_p5mr_exposure`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_x746_7m8f_x49c_fix_pinning`	cve	fail	fail	blocked on both versions: RuntimeError on old and new; possible package or environment defect, reproducer attached	open issue; reproducer is the test; advisory stays under investigation
`test_ghsa_x746_7m8f_x49c_exposure`	cve	fail	fail	blocked on both versions: RuntimeError on old and new; possible package or environment defect, reproducer attached	open issue; reproducer is the test; advisory stays under investigation
`test_gdsa_wqp7_x3pw_xc5r_fix_pinning`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_gdsa_wqp7_x3pw_xc5r_exposure`	cve	pass	pass	not a fix-pinning test: passes on both versions; keep only as characterization if it covers the symbol	discard as CVE evidence; keep as characterization only if reviewer wants it
`test_ghsa_jp82_jpqv_5vv3_fix_pinning`	cve	fail	fail	fails on the fixed version: test bug or the advisory is misread; back to generation	regenerate or discard
`test_ghsa_jp82_jpqv_5vv3_exposure`	cve	fail	fail	fails on the fixed version: test bug or the advisory is misread; back to generation	regenerate or discard
`test_ghsa_82w8_qh3p_5jfq_fix_pinning`	cve	pass	fail	exposure confirmed: fails at head 0.41.3, passes with the fixed candidate 1.3.1	accept as CVE evidence; VEX status affected: the change introduces the exposure
`test_ghsa_82w8_qh3p_5jfq_exposure`	cve	pass	fail	exposure confirmed: fails at head 0.41.3, passes with the fixed candidate 1.3.1	accept as CVE evidence; VEX status affected: the change introduces the exposure
`test_simple_user_authenticated`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_cookie_parser_basic`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_response_status_and_body`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate
`test_json_response_renders_json`	unit	pass	pass	candidate: passes on head, same on old	accept as characterization candidate

Findings

Class	Confidence	Finding	Route
security	0.9	starlette 0.41.3 has 14 open advisories at head; reachable=true	Product Security
defect	0.4	starlette: a code path raised the same internal error on both versions during CVE test generation (CVE-2025-54121); NOT corroborated by stage 4 verdicts	escalate: below confidence threshold
defect	0.4	starlette: a code path raised the same internal error on both versions during CVE test generation (CVE-2026-54283); NOT corroborated by stage 4 verdicts	escalate: below confidence threshold
defect	0.8	starlette: a code path raised the same internal error on both versions during CVE test generation (CVE-2026-48817)	open issue with reproducer
obsolete	0.85	starlette: test_ghsa_jp82_jpqv_5vv3_exposure references starlette.requests.Request, which the upgrade to 1.3.1 changes the required parameters of	propose retirement
redundant	0.5	starlette: test_simple_user_authenticated killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	starlette: test_cookie_parser_basic killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	starlette: test_response_status_and_body killed none of 12 sampled mutants in this run	escalate: below confidence threshold
redundant	0.5	starlette: test_json_response_renders_json killed none of 12 sampled mutants in this run	escalate: below confidence threshold

Draft VEX statements, for Product Security

Vulnerability	Status	Basis
CVE-2025-54121	affected	open at head; reachable=true; fixed in ['0.47.2']
CVE-2025-62727	affected	open at head; reachable=true; fixed in ['0.49.1']
CVE-2026-54283	affected	exposure test(s) ['test_ghsa_82w8_qh3p_5jfq_exposure', 'test_ghsa_82w8_qh3p_5jfq_fix_pinning'] fail on pkg:pypi/starlette@0.41.3 and pass on pkg:pypi/starlette@0.41.3: the change i
CVE-2026-48710	affected	open at head; reachable=true; fixed in ['1.0.1']
CVE-2026-54282	affected	open at head; reachable=true; fixed in ['1.3.0']
CVE-2026-48818	affected	open at head; reachable=true; fixed in ['1.1.0']
CVE-2026-48817	affected	open at head; reachable=true; fixed in ['1.1.0']

Retirement proposals

Test	Reason	Evidence
`test_ghsa_jp82_jpqv_5vv3_exposure`	obsolete-symbol-removed	test_ghsa_jp82_jpqv_5vv3_exposure references starlette.requests.Request, which the upgrade to 1.3.1 changes the required parameters of
`test_simple_user_authenticated`	redundant-no-kills	test_simple_user_authenticated killed none of 12 sampled mutants in this run
`test_cookie_parser_basic`	redundant-no-kills	test_cookie_parser_basic killed none of 12 sampled mutants in this run
`test_response_status_and_body`	redundant-no-kills	test_response_status_and_body killed none of 12 sampled mutants in this run
`test_json_response_renders_json`	redundant-no-kills	test_json_response_renders_json killed none of 12 sampled mutants in this run

The records, in plain terms

60 UDLM records, each a fact from this run in the estate's own language. Intent is what the harness proposed, Requested is what a person was asked to decide, Realized is what landed and who made it so. All are sealed with a content hash.

Record	State	What it says	Written
Job	Requested	The harness was asked to run on frc-scheduler-server from 2ad04243861b to cfa8f4a3292b in diff mode.	2026-09-12 01:41:41 UTC
Job	Realized	The run finished; its outputs are the records this statement seals.	2026-09-13 02:40:04 UTC
SoftwarePackage	Discovered	pyasn1 0.4.8 (pypi) is the version this change installs.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_debug_is_callable` is a unit test in `tests/test_pyasn1_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_ghsa_jr27_m4p2_rc6r_exposure` is an exposure test for CVE-2026-30922 in `tests/test_pyasn1_cve_cve_2026_30922.py`; it failed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_ghsa_jr27_m4p2_rc6r_fix_pinning` is a fix-pinning test for CVE-2026-30922 in `tests/test_pyasn1_cve_cve_2026_30922.py`; it failed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_hexdump_empty_bytes` is a unit test in `tests/test_pyasn1_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_printer_is_callable` is a unit test in `tests/test_pyasn1_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_pyasn1error_is_exception_subclass` is a unit test in `tests/test_pyasn1_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_pyasn1unicode_error_is_unicode_error_subclass` is a unit test in `tests/test_pyasn1_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_setlogger_returns_none` is a unit test in `tests/test_pyasn1_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_substrate_underrun_error_is_pyasn1error_subclass` is a unit test in `tests/test_pyasn1_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
TestEvidence	Intent	`test_value_constraint_error_is_pyasn1error_subclass` is a unit test in `tests/test_pyasn1_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 02:40:04 UTC
VexStatement	Intent	Draft: CVE-2026-59885 is `under_investigation` in this package. open at head; reachable=unknown; fixed in ['0.6.4']	2026-09-13 02:40:04 UTC
VexStatement	Intent	Draft: CVE-2026-59886 is `under_investigation` in this package. open at head; reachable=unknown; fixed in ['0.6.4']	2026-09-13 02:40:04 UTC
VexStatement	Intent	Draft: CVE-2026-30922 is `affected` in this package. exposure test(s) ['test_ghsa_jr27_m4p2_rc6r_fix_pinning', 'test_ghsa_jr27_m4p2_rc6r_exposure'] fail on pkg:pypi/pyasn1@0.4.8 and pass on pkg:pypi/pyasn1@0.6.4: the change introduces this exposure	2026-09-13 02:40:04 UTC
VexStatement	Intent	Draft: CVE-2026-59884 is `under_investigation` in this package. open at head; reachable=unknown; fixed in ['0.6.4']	2026-09-13 02:40:04 UTC
Vulnerability	Discovered	CVE-2026-59885 is a known vulnerability affecting <0.6.4. Also known as GHSA-8ppf-4f7h-5ppj, PYSEC-2026-3456.	2026-09-13 02:40:04 UTC
Vulnerability	Discovered	CVE-2026-59886 is a known vulnerability affecting <0.6.4. Also known as GHSA-hm4w-wwcw-mr6r, PYSEC-2026-3457.	2026-09-13 02:40:04 UTC
Vulnerability	Discovered	CVE-2026-30922 is a known vulnerability affecting <0.6.3. Also known as GHSA-jr27-m4p2-rc6r, PYSEC-2026-2263.	2026-09-13 02:40:04 UTC
Vulnerability	Discovered	CVE-2026-59884 is a known vulnerability affecting <0.6.4. Also known as GHSA-m4p7-r5rc-7g4j, PYSEC-2026-3455.	2026-09-13 02:40:04 UTC
Job	Requested	The harness was asked to run on frc-scheduler-server from 2ad04243861b to cfa8f4a3292b in diff mode.	2026-09-12 01:41:41 UTC
Job	Realized	The run finished; its outputs are the records this statement seals.	2026-09-12 06:51:29 UTC
SoftwarePackage	Discovered	python-jose 3.4.0 (pypi) is the version this change installs.	2026-09-12 06:51:29 UTC
TestEvidence	Intent	`test_ghsa_cjwg_qfpm_7377_exposure` is an exposure test for CVE-2024-33664 in `tests/test_python_jose_cve_cve_2024_33664.py`; it passed on head in the sealed sandbox.	2026-09-12 06:51:29 UTC
TestEvidence	Intent	`test_ghsa_cjwg_qfpm_7377_fix_pinning` is a fix-pinning test for CVE-2024-33664 in `tests/test_python_jose_cve_cve_2024_33664.py`; it passed on head in the sealed sandbox.	2026-09-12 06:51:29 UTC
TestEvidence	Intent	`test_jwt_decode_expired_raises_expiredsignatureerror` is a unit test in `tests/test_python_jose_unit.py`; it passed on head in the sealed sandbox.	2026-09-12 06:51:29 UTC
TestEvidence	Intent	`test_jwt_decode_wrong_key_raises_jwterror` is a unit test in `tests/test_python_jose_unit.py`; it passed on head in the sealed sandbox.	2026-09-12 06:51:29 UTC
TestEvidence	Intent	`test_jwt_encode_decode_roundtrip` is a unit test in `tests/test_python_jose_unit.py`; it passed on head in the sealed sandbox.	2026-09-12 06:51:29 UTC
TestEvidence	Intent	`test_jwt_get_unverified_claims` is a unit test in `tests/test_python_jose_unit.py`; it passed on head in the sealed sandbox.	2026-09-12 06:51:29 UTC
TestEvidence	Intent	`test_jwt_get_unverified_header` is a unit test in `tests/test_python_jose_unit.py`; it passed on head in the sealed sandbox.	2026-09-12 06:51:29 UTC
VexStatement	Intent	Draft: CVE-2024-33663 is `under_investigation` in this package. bump to a fixed version per advisory metadata; no confirmed fix-pinning test yet	2026-09-12 06:51:29 UTC
VexStatement	Intent	Draft: CVE-2024-33664 is `fixed` in this package. fix-pinning test(s) ['test_ghsa_cjwg_qfpm_7377_exposure', 'test_ghsa_cjwg_qfpm_7377_fix_pinning'] fail on pkg:pypi/python-jose@3.3.0 and pass on pkg:pypi/python-jose@3.4.0	2026-09-12 06:51:29 UTC
VexStatement	Intent	Draft: CVE-2024-29370 is `under_investigation` in this package. bump to a fixed version per advisory metadata; no confirmed fix-pinning test yet	2026-09-12 06:51:29 UTC
Vulnerability	Discovered	CVE-2024-33663 is a known vulnerability affecting <3.4.0. Also known as GHSA-6c5p-j8vq-pqhj, PYSEC-2024-232.	2026-09-12 06:51:29 UTC
Vulnerability	Discovered	CVE-2024-33664 is a known vulnerability affecting <3.4.0. Also known as GHSA-cjwg-qfpm-7377, PYSEC-2024-233.	2026-09-12 06:51:29 UTC
Vulnerability	Discovered	CVE-2024-29370 is a known vulnerability affecting versions the advisory does not pin. Also known as PYSEC-2025-185.	2026-09-12 06:51:29 UTC
Job	Requested	The harness was asked to run on frc-scheduler-server from 2ad04243861b to cfa8f4a3292b in diff mode.	2026-09-12 01:41:41 UTC
Job	Realized	The run finished; its outputs are the records this statement seals.	2026-09-13 04:29:47 UTC
SoftwarePackage	Discovered	starlette 0.41.3 (pypi) is the version this change installs.	2026-09-13 04:29:47 UTC
TestEvidence	Intent	`test_cookie_parser_basic` is a unit test in `tests/test_starlette_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 04:29:47 UTC
TestEvidence	Intent	`test_ghsa_82w8_qh3p_5jfq_exposure` is an exposure test for CVE-2026-54283 in `tests/test_starlette_cve_cve_2026_54283.py`; it failed on head in the sealed sandbox.	2026-09-13 04:29:47 UTC
TestEvidence	Intent	`test_ghsa_82w8_qh3p_5jfq_fix_pinning` is a fix-pinning test for CVE-2026-54283 in `tests/test_starlette_cve_cve_2026_54283.py`; it failed on head in the sealed sandbox.	2026-09-13 04:29:47 UTC
TestEvidence	Intent	`test_json_response_renders_json` is a unit test in `tests/test_starlette_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 04:29:47 UTC
TestEvidence	Intent	`test_response_status_and_body` is a unit test in `tests/test_starlette_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 04:29:47 UTC
TestEvidence	Intent	`test_simple_user_authenticated` is a unit test in `tests/test_starlette_unit.py`; it passed on head in the sealed sandbox.	2026-09-13 04:29:47 UTC
VexStatement	Intent	Draft: CVE-2025-54121 is `affected` in this package. open at head; reachable=true; fixed in ['0.47.2']	2026-09-13 04:29:47 UTC
VexStatement	Intent	Draft: CVE-2025-62727 is `affected` in this package. open at head; reachable=true; fixed in ['0.49.1']	2026-09-13 04:29:47 UTC
VexStatement	Intent	Draft: CVE-2026-54283 is `affected` in this package. exposure test(s) ['test_ghsa_82w8_qh3p_5jfq_exposure', 'test_ghsa_82w8_qh3p_5jfq_fix_pinning'] fail on pkg:pypi/starlette@0.41.3 and pass on pkg:pypi/starlette@0.41.3: the change introduces this exposure	2026-09-13 04:29:47 UTC
VexStatement	Intent	Draft: CVE-2026-48710 is `affected` in this package. open at head; reachable=true; fixed in ['1.0.1']	2026-09-13 04:29:47 UTC
VexStatement	Intent	Draft: CVE-2026-54282 is `affected` in this package. open at head; reachable=true; fixed in ['1.3.0']	2026-09-13 04:29:47 UTC
VexStatement	Intent	Draft: CVE-2026-48818 is `affected` in this package. open at head; reachable=true; fixed in ['1.1.0']	2026-09-13 04:29:47 UTC
VexStatement	Intent	Draft: CVE-2026-48817 is `affected` in this package. open at head; reachable=true; fixed in ['1.1.0']	2026-09-13 04:29:47 UTC
Vulnerability	Discovered	CVE-2025-54121 is a known vulnerability affecting <0.47.2. Also known as GHSA-2c2j-9gv5-cj73, PYSEC-2026-1941.	2026-09-13 04:29:47 UTC
Vulnerability	Discovered	CVE-2025-62727 is a known vulnerability affecting <0.49.1. Also known as GHSA-7f5h-v6xp-fcq8, PYSEC-2026-1942.	2026-09-13 04:29:47 UTC
Vulnerability	Discovered	CVE-2026-54283 is a known vulnerability affecting <1.3.1. Also known as GHSA-82w8-qh3p-5jfq, PYSEC-2026-249.	2026-09-13 04:29:47 UTC
Vulnerability	Discovered	CVE-2026-48710 is a known vulnerability affecting <1.0.1. Also known as GHSA-86qp-5c8j-p5mr, PYSEC-2026-161.	2026-09-13 04:29:47 UTC
Vulnerability	Discovered	CVE-2026-54282 is a known vulnerability affecting <1.3.0. Also known as GHSA-jp82-jpqv-5vv3, PYSEC-2026-248.	2026-09-13 04:29:47 UTC
Vulnerability	Discovered	CVE-2026-48818 is a known vulnerability affecting <1.1.0. Also known as GHSA-wqp7-x3pw-xc5r, PYSEC-2026-2281.	2026-09-13 04:29:47 UTC
Vulnerability	Discovered	CVE-2026-48817 is a known vulnerability affecting <1.1.0. Also known as GHSA-x746-7m8f-x49c, PYSEC-2026-2280.	2026-09-13 04:29:47 UTC

The run against the blueprint's goals

Id	Goal	Measured	Verdict	From
G1	Nothing merges without a person	harness produced patches and drafts only; no git push, no VEX published	met	`packet.md` `packet.md`
G2	Every kept test was compiled and run in the sandbox before a reviewer sees it	45 tests ran on head, again for flakes, and on base; 0 flaky	met	`results.json` `results.json`
G3	A test that passes on both versions is not presented as CVE evidence	6 fix-pinning confirmed; pass-on-both CVE tests are classified test-bug and excluded from VEX 'fixed'	met	`triage.json` `triage.json`
G4	Stage 0 passed before evidence was produced	28 of 28 register checks; sandbox probe skipped; model probe skipped	met	`selfcheck.json`
G5	Generated tests ran with no network, no secrets, read-only root, disposable container	{'capabilities': 'all dropped', 'disposable': True, 'limits': {'cpus': '2', 'memory': '2g', 'pids': 512, 'timeout_s': 90	met	`sandbox.json` `sandbox.json`
G6	Every accepted test carries a provenance record and a signed attestation	23 records; envelopes verified locally: [True, True, True]; signer is a development key	met (development signer)	`statement.dsse.json` `statement.dsse.json`
G7	Every advisory on the work item has a CVE-targeted test attempt and a draft VEX statement	pyasn1: {'CVE-2026-30922': 'affected', 'CVE-2026-59884': 'under_investigation', 'CVE-2026-59885': 'under_investigation', 'CVE-2026-59886': 'under	met	`vex.openvex.json` `vex.openvex.json`
G8	The run surfaces a finding the PR diff does not show	pyasn1: ['pyasn1 downgraded 0.6.4 -> 0.4.8 into a version with 8 known advisories', 'pyasn1 0.4.8 has 8 open advisories at head; reachable=unknown', 'pyasn1: a code path raised the same intern	met	`triage.json` `triage.json`
G9	Packet within 2 hours of trigger for depth 0 and 1	intake 2026-09-12T01:42:29+00:00 -> packet 2026-09-13T16:25:41+00:00	see measurement	`run.json` `summary.json`
G10	Every number in the packet points at a file in the work directory	packet sections cite artifact paths; attestation subjects are the patch and the manifest digests	met	`statement.json` `statement.json`
G11	Test strength measured by mutation and coverage delta	pyasn1: score 0.417 on 12 sampled mutants; python-jose: score 0.417 on 12 sampled mutants; starlette: score 0.0 on 12 sampled mutants; coverage delta: no baseline overlay suite yet	not met (target 0.6 on sampled mutants)	`mutation.json` `mutation.json`
G12	The pipeline's own gaps are stated, never inherited silently	analysis predates the pipeline facts; rerun analysis	not applicable
G13	Every behaviour the change reaches ends in exactly one state, and a gap carries its reason	5 behaviour(s) reached: 0 validated, 0 for your environment, 5 stated gap(s) (5 strength-not-sampled); 2 reference(s) set aside with a reason	met	`accounting-summary.json` `accounting.json`

What is in this folder, and who it is for

Every file exists for a reader or a tool named here. The plain-English files come first; the machine files stay because they are the proof and the replay.

Everyone

README.md: The story of this run: who, what, why, where, when, the outcome, the decisions and the actions.
triage/accounting-summary.json: The accounting rolled up: how many behaviours were reached, validated, handed to the team, or left as stated gaps, and why.
assess/assess.md: The run against the blueprint's thirteen goals, each with the measurement and the file it came from.

Reviewers (the developer or team that owns the change)

Product Security

intake/vulns.json: Known vulnerabilities for every package version in either graph, from OSV, including malicious-package reports.
analyze/pyasn1/vulns.json: The advisories on this package's versions, with the symbols they name and the fixed versions.
analyze/python-jose/vulns.json: The advisories on this package's versions, with the symbols they name and the fixed versions.
analyze/starlette/vulns.json: The advisories on this package's versions, with the symbols they name and the fixed versions.
analyze/starlette/fixed-candidate.json: When the change left a vulnerable version in place: the environment the harness resolved with the fixed version, and what had to move.
packet/pyasn1/vex.openvex.json: One draft VEX statement per advisory, with its basis; a proposal for Product Security, never published by the harness.
packet/python-jose/vex.openvex.json: One draft VEX statement per advisory, with its basis; a proposal for Product Security, never published by the harness.
packet/starlette/vex.openvex.json: One draft VEX statement per advisory, with its basis; a proposal for Product Security, never published by the harness.

Supply Chain Security and auditors

The person who ran the harness

analyze/pyasn1/source-diff.patch: The package's own source diff between the versions; the model reads it to find the fix.
analyze/python-jose/source-diff.patch: The package's own source diff between the versions; the model reads it to find the fix.
analyze/pyasn1/notes.untrusted.md: Package metadata and description as published upstream; untrusted text, kept for context.
analyze/python-jose/notes.untrusted.md: Package metadata and description as published upstream; untrusted text, kept for context.
analyze/starlette/notes.untrusted.md: Package metadata and description as published upstream; untrusted text, kept for context.
generate/pyasn1/manifest.json: What was generated: every file, every test, what was cut and discarded and why, the model, the prompt set and its digest, the prompt and response digests.
generate/python-jose/manifest.json: What was generated: every file, every test, what was cut and discarded and why, the model, the prompt set and its digest, the prompt and response digests.
generate/starlette/manifest.json: What was generated: every file, every test, what was cut and discarded and why, the model, the prompt set and its digest, the prompt and response digests.
generate/pyasn1/manifest.agent.json: The tool-using agent's traces: every tool call per advisory, and the budget it had.
generate/python-jose/manifest.agent.json: The tool-using agent's traces: every tool call per advisory, and the budget it had.
generate/starlette/manifest.agent.json: The tool-using agent's traces: every tool call per advisory, and the budget it had.

Tools (kept for replay and proof; not meant to be read)

Rendered by tools/build-runs.py from run-index.json and the files it names, with the story from harness/src/harness/story.py. Every value is escaped; nothing here is executed or trusted.

Who, what, why, where, when

Decisions the harness made

Actions it took

Actions it did not take, by design

How it happened, stage by stage

Read

Did

Wrote

Read

Did

Wrote

Read

Did

Wrote

Read

Did

Wrote

Read

Did

Wrote

Read

Did

Wrote

Read

Did

Wrote

Read

Did

Wrote

Read

Did

Wrote

Read

Did

Wrote

Read

Did

Wrote

pyasn1 0.6.4 → 0.4.8

Tests, one row each

Findings

Draft VEX statements, for Product Security

Mutation: which test killed what

Retirement proposals

python-jose 3.3.0 → 3.4.0

Tests, one row each

Findings

Draft VEX statements, for Product Security

Mutation: which test killed what

Retirement proposals

starlette 0.41.3 0.41.3

Tests, one row each

Findings

Draft VEX statements, for Product Security

Retirement proposals

The records, in plain terms

The run against the blueprint's goals

What is in this folder, and who it is for

What is in this folder, and who it is for

Everyone

Reviewers (the developer or team that owns the change)

Product Security

Supply Chain Security and auditors

The person who ran the harness

Tools (kept for replay and proof; not meant to be read)