# Captured by scripts/generate-proof.mjs
# node v24.15.0 on win32/x64 (10.0.26300)
# exit code: 0
# This file is the command output. It is not a summary written by hand.

✔ doctor succeeds without a provider key (1996.9432ms)
✔ agents catalog includes Prime and 16 specialists (9.053ms)
✔ focused test is high-severity blocker with exact evidence (33.1791ms)
✔ skipped test is reported and not disguised as executed test (27.5559ms)
✔ no tests, CI or license produce transparent warnings (12.715ms)
✔ clean static project still requires review without test run (60.5495ms)
✔ report HTML escapes untrusted strings (22.184ms)
✔ scan command creates three reports without executing tests (1523.4413ms)
✔ release-check exit code signals NO_GO (1664.2472ms)
✔ release-check exit code signals REVIEW_REQUIRED (991.3082ms)
✔ installer copies canonical skills into Claude target and prevents overwrite (687.6993ms)
✔ installer refuses a target file or missing directory (16.5616ms)
✔ unknown options or missing values fail clearly (2635.3921ms)
✔ nested symlink is skipped rather than followed (16.0673ms)
✔ heuristic avoids matches embedded in fixture strings (34.5388ms)
✔ reports do not disclose absolute target paths by default (11.3816ms)
✔ passing CI evidence on clean project reaches READY_FOR_HUMAN_APPROVAL (exit 0) (1846.559ms)
✔ failing evidence yields NO_GO (1229.3771ms)
✔ external-unverified evidence never reaches READY (936.2362ms)
✔ zero-pass evidence is inconclusive, absent evidence is never READY (2972.4728ms)
✔ verify rejects malformed evidence and accepts valid evidence (4124.4385ms)
✔ version and --version print the package version (2372.5482ms)
✔ release report quotes supplied evidence in Markdown and HTML (793.3989ms)
✔ pytest.skip(), test.todo and binary tests are handled (22.3421ms)
✔ pytest/JUnit/Go skip markers are reported (30.1757ms)
✔ missing path and traversal-style output fail closed without stack trace (665.6004ms)
✔ HTML and Markdown reports escape malicious file names (11.7642ms)
✔ scan writes md, html and json reports by default into ./pramanasutra-reports (731.3379ms)
✔ --no-report writes nothing (691.0859ms)
✔ scan rejects --evidence/--policy; they are release-check only (618.9921ms)
✔ strict policy turns skipped tests into NO_GO; recommended does not (1253.3661ms)
✔ policy is fail-closed: unknown keys and weakened safety flags are rejected (1804.4862ms)
✔ bundled policies load (1163.9838ms)
✔ Java/C#/Ruby naming conventions are detected, look-alikes are not (11.9651ms)
✔ report markdown and html contain decision, evidence location and remediation (8.098ms)
✔ committed fixtures produce the expected rules and decisions (6.7109ms)
✔ .pramanasutraignore excludes labeled fixtures and rejects traversal (16.7889ms)
✔ skills: unique names, frontmatter matches folder, non-empty workflow, no network or execution instructions (9.2608ms)
✔ CircleCI config is recognised as CI (16.7453ms)
✔ evidence binding is quoted for unbound and ci-environment files (0.6379ms)
✔ SARIF 2.1.0 is written beside the other reports (596.9598ms)
✔ --changed without a git repository says diff was unavailable (728.6906ms)
✔ help says scan does not fail the build (591.4412ms)
ℹ tests 43
ℹ suites 0
ℹ pass 43
ℹ fail 0
ℹ cancelled 0
ℹ skipped 0
ℹ todo 0
ℹ duration_ms 33972.5163