chore(release): bump llm-security to v7.1.0

Closes A4 of v7.1.0 critical-review patch — release artefacts.

- Version bump 7.0.0 → 7.1.0 across active version sources:
  * package.json
  * .claude-plugin/plugin.json
  * CLAUDE.md header
  * README.md badge
  * scanners/ide-extension-scanner.mjs (VERSION constant)
  * marketplace root README plugin entry
- Marketplace root README test count: 1487 → 1511.
- CHANGELOG.md: new [7.1.0] - 2026-04-29 section above [7.0.0],
  documenting B1, B2, B4, B8, honesty-sweep (7 phrases), and
  test-count delta (+24 → 1511 total).
- docs/security-hardening-guide.md: §6 last-updated bump + new
  v7.1.0 calibration note on hook-level fixes (pathguard regex
  hole, distributed-trifecta block-mode bypass).

Historical references to "7.0.0" intentionally preserved in:
- CHANGELOG [7.0.0] entries (history)
- README.md version-history table v5.0.0/v7.0.0 rows (history)
- CLAUDE.md §"v7.0.0 — Severity-dominated risk scoring" (describes
  what changed at v7.0.0 release)
- scanners/ JSDoc comments noting "v7.0.0+" formula provenance
- agents/ + tests/ + knowledge/ provenance comments

Pre-existing untracked/modified tracker noise (.gitignore,
marketplace.json, config-audit/docs, ultraplan-local/docs) is not
part of this commit per the v7.1.0 NEXT-SESSION-PROMPT handoff.

Tests: 1511/1511 green.
This commit is contained in:
Kjell Tore Guttormsen 2026-04-29 11:57:16 +02:00
commit 621db144bd
8 changed files with 65 additions and 8 deletions

View file

@ -269,4 +269,19 @@ tools.
---
**Last updated:** 2026-04-19 for v7.0.0.
**Last updated:** 2026-04-29 for v7.1.0.
### v7.1.0 calibration note
v7.1.0 is a patch release. No calibration changes; the §6 tuning workflow above is
unchanged. Two hook-level bugs were fixed that affect production posture:
- `pre-write-pathguard.mjs` now blocks multi-segment `.env.*.*.*` paths (previously a
regex hole let `.env.production.local.backup` through).
- `post-session-guard.mjs` `block` mode now blocks every detected trifecta. Previously
required a "concentrated MCP" or "sensitive path" qualifier, so distributed
trifectas were advisory-only even in block mode.
If you run with `LLM_SECURITY_TRIFECTA_MODE=block`, expect the false-block rate to
rise after this upgrade — the previous gate suppressed real trifectas. Re-baseline
the warn-mode noise floor before promoting to block, per §3.