fix(agent-sec-core): Refactor SKILL.md to executable protocol and align sub-skills#130
Merged
kid9 merged 3 commits intoalibaba:mainfrom Apr 9, 2026
Merged
Conversation
…utable protocol - Reorganize by LLM execution order: priority declaration, hard block rule definitions, Phase 1-3 workflow, security decision, self-protection - Add output anchoring: forced first-step status table, fixed output format per Phase (PASS/FAIL/NOT_RUN), self-terminating block template - Replace declarative constraints with imperative instructions to prevent Phase skipping and non-execution after trigger - Eliminate rule conflicts: split into absolute prohibition and high-risk-confirmation tiers with no overlap - Hard block rules defined in Section 2, applied in Section 4 security decision judgment chain step 2 - Add Decision UNLOCKED state in Phase 3 PASS output for state machine symmetry - Remove threat model table, audit log format (moved to README), exception handling table, Gate A/B/C naming, broken hardening-checklist.md reference BREAKING CHANGE: SKILL.md structure completely rewritten from 5-principle policy manual to 5-section executable protocol
agent-sec-seharden.md (Phase 1): - Add phase: 1 to frontmatter - Add Status Line Output section: [Phase 1] PASS/FAIL/NOT_RUN agent-sec-skill-verify.md (Phase 2): - Add phase: 2 to frontmatter - Add prerequisites section (gpg/gnupg2 and verifier.py checks) - Add Status Line Output section: [Phase 2] PASS/FAIL/NOT_RUN
- Update architecture diagram: show security check workflow and sandbox policy as separate channels - Add PASS condition column to Phase 1-3 workflow table - Add Audit Log section (format moved from SKILL.md) - Update SKILL.md description to executable security protocol - Sync README_CN.md with English version
|
|
edonyzpc
changed the title
1570005763
approved these changes
Apr 8, 2026
Collaborator
1570005763
left a comment
1570005763
left a comment
There was a problem hiding this comment.
LGTM.
Heads up — agent-sec-core is also blocking legitimate skill changes, which is kind of expected for now since we're being conservative on the security side. But going forward, we'll probably need to think about finer-grained risk management around skill signing — like, being able to tell apart a normal incremental update from something actually suspicious, instead of flagging everything the same way.
Collaborator
|
LGTM. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
Related Issue
closes #
Type of Change
Scope
cosh(copilot-shell)sec-core(agent-sec-core)skill(os-skills)sight(agentsight)Checklist
cosh: Lint passes, type check passes, and tests passsec-core(Rust):cargo clippy -- -D warningsandcargo fmt --checkpasssec-core(Python): Ruff format and pytest passskill: Skill directory structure is valid and shell scripts pass syntax checksight:cargo clippy -- -D warningsandcargo fmt --checkpasspackage-lock.json/Cargo.lock)Testing
Additional Notes