docs: record AgentShield plugin-cache evidence

This commit is contained in:
Affaan Mustafa
2026-05-16 00:27:48 -04:00
parent 6c8e909d63
commit 1eb7b0809d
7 changed files with 43 additions and 19 deletions

View File

@@ -1,6 +1,6 @@
# AgentShield Enterprise Research Roadmap
Generated: 2026-05-12
Generated: 2026-05-12; refreshed with May 16 AgentShield PR #87 evidence.
This is a planning artifact for the next AgentShield enterprise iteration. It
does not modify AgentShield code. The goal is to turn the current scanner,
@@ -84,8 +84,15 @@ AgentShield is already more than a static lint tool:
- Enterprise hooks exist: policy packs, exception metadata, expiring/expired
exception reporting, SARIF code scanning, and job-summary output.
- Accuracy work is active: `runtimeConfidence`, template/example weighting,
docs-example downgrades, hook-manifest resolution, false-positive audit
guidance, and corpus readiness.
docs-example downgrades, installed Claude plugin-cache confidence,
hook-manifest resolution, false-positive audit guidance, and corpus readiness.
May 16 update: AgentShield PR #87 merged as
`26bb44650663816d07180e0d20c1895e431a326c`. It classifies installed Claude
plugin cache content as `runtimeConfidence: plugin-cache`, keeps non-secret
plugin-cache score impact at `0.5x`, avoids downgrading repository-local
non-Claude `plugins/cache` paths, and makes plugin-cache classification win
before cached hook implementations would otherwise appear as active `hook-code`.
The next iteration should not be "add more regex rules" by default. The higher
leverage move is to make AgentShield remember, compare, route, and enforce
@@ -323,6 +330,8 @@ The AgentShield enterprise iteration is not complete until these are true:
- Built CLI smoke tests cover the new flags or report modes.
- GitHub Action self-test covers the new CI-visible output.
- Documentation names the free/local path and the paid/team path separately.
- Runtime-confidence changes include live scan evidence proving lower-confidence
plugin/package surfaces stay visible instead of being suppressed.
- Evidence produced by the feature is deterministic enough for CI diffing.
- ECC-Tools can consume the finding fingerprints or backlog export without
exceeding GitHub/Linear object caps.