You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@magpie.apache.org by "potiuk (via GitHub)" <gi...@apache.org> on 2026/09/27 20:58:18 UTC

[PR] perf(security): trim the always-on descriptions of the security-model skills [magpie]

potiuk opened a new pull request, #1438:
URL: https://github.com/apache/magpie/pull/1438

   Fourth of a four-PR stack for the **security** family (#1344).
   
   ## Summary
   
   - `description` + `when_to_use` load in every session, for every installed skill. The three `security-model-*` skills spent ~250–323 tokens there, against the 200-token budget. Most of it was rationale the body already covers: the Alpha-Omega delegation, the draft-first reasoning, the regression-check motive, and routing clauses repeated across all three skills.
   - Each skill keeps its purpose, its main trigger phrases, and the one neighbouring skill to route to instead.
   
   | skill | always-on (chars ÷ 4) |
   |---|---:|
   | security-model-update | ~323 → ~122 |
   | security-model-verify | ~293 → ~125 |
   | security-model-prepare | ~251 → ~110 |
   
   `docs/setup/marketplace.md`'s always-on figure for `magpie-security` is updated from ~1.9k to ~1.6k, as measured by `check-doc-sync`.
   
   Other security skills are still over the always-on budget: `issue-import-via-forwarder` (~291), `issue-import` (~280), `issue-triage` (~278), `issue-invalidate` (~251), `issue-import-from-pr` (~239), `issue-import-from-scan` (~237), `issue-fix` (~214) and `cve-allocate` (~203). They go in a follow-up wording pass.
   
   ## Type of change
   
   - [x] Skill change: frontmatter only
   - [x] Documentation: `docs/setup/marketplace.md`
   
   ## Test plan
   
   - [x] `prek run --all-files` passes
   - [x] `skill-and-tool-validate` green
   - [x] Frontmatter is not part of any eval's extracted step section, so eval inputs are unchanged
   
   ## Stack
   
   1. drop Snapshot drift
   2. split into siblings
   3. one safe-create recipe, board-recipe pointers
   4. **this PR**: trim the `security-model-*` always-on descriptions
   
   Part of #1344.
   
   Generated-by: Claude Code (Opus 5.5)
   
   🤖 Generated with [Claude Code](https://claude.com/claude-code)
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@magpie.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


Re: [PR] perf(security): trim the always-on descriptions of the security-model skills [magpie]

Posted by "potiuk (via GitHub)" <gi...@apache.org>.
potiuk merged PR #1438:
URL: https://github.com/apache/magpie/pull/1438


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@magpie.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org