Skip to content

Dev - #479

Merged
alirezarezvani merged 6 commits into
mainfrom
dev
Apr 3, 2026
Merged

Dev#479
alirezarezvani merged 6 commits into
mainfrom
dev

Conversation

@alirezarezvani

Copy link
Copy Markdown
Owner

Summary

Checklist

  • Target branch is dev (not main — PRs to main will be auto-closed)
  • Skill has SKILL.md with valid YAML frontmatter (name, description, license)
  • Scripts (if any) run with --help without errors
  • No hardcoded API keys, tokens, or secrets
  • No vendor-locked dependencies without open-source fallback
  • Follows existing directory structure (domain/skill-name/SKILL.md)

Type of Change

  • New skill
  • Improvement to existing skill
  • Bug fix
  • Documentation
  • Infrastructure / CI

Testing

alirezarezvani and others added 6 commits April 3, 2026 01:54
Based on issue #464 submission by voidborne-d. Enhanced with English-only
content (removed all Chinese), anti-patterns section, cross-references,
plugin.json, convention-compliant frontmatter, and English eval scenarios.

behuman (193 lines + reference + 8 eval scenarios):
- Self-Mirror loop: instinctive response → reflection → conscious revision
- Show mode (2.5-3x tokens) and quiet mode (1.5-2x tokens)
- 3 English examples: emotional support, life advice, personal writing
- Based on Lacan's Mirror Stage + Kahneman's Dual Process Theory
- Zero dependencies — pure prompt technique

Co-Authored-By: voidborne-d <voidborne-d@users.noreply.github.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
feat(engineering): add behuman skill — Self-Mirror consciousness loop
Add a skill for creating CodeTour .tour files — persona-targeted,
step-by-step walkthroughs that link to real files and line numbers.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
feat(engineering): add code-tour skill
@alirezarezvani
alirezarezvani merged commit caf17d6 into main Apr 3, 2026
4 of 5 checks passed
@claude

claude Bot commented Apr 3, 2026

Copy link
Copy Markdown

PR Review — #479 (Dev → Main)

Note: This PR is already merged. The following is a post-merge audit for tracking purposes and future reference.


Summary

This PR adds two new engineering-POWERFUL skills:

  • behuman — Self-Mirror Consciousness Loop for more human-like AI responses
  • code-tour — CodeTour .tour file generator for VS Code

Both are solid concepts. behuman in particular is well-executed with good examples, an evals file, and clear anti-patterns. code-tour provides a useful structured workflow. A few issues noted below.


Issues Found

🔴 Critical

1. SKILL.md frontmatter missing license field (both skills)

Per the PR checklist and CLAUDE.md, SKILL.md must have name, description, and license in YAML frontmatter.

engineering/behuman/SKILL.md:

---
name: "behuman"
description: "..."
# ❌ license field missing
---

engineering/code-tour/SKILL.md:

---
name: "code-tour"
description: "..."
# ❌ license field missing
---

Fix: add license: "MIT" to both.


🟡 Medium

2. docs/skills/engineering/index.md count is off

Two skills were added (behuman + code-tour) but the count only went from 51 → 52. Should be 53.

-<p class="domain-count">52 skills in this domain</p>
+<p class="domain-count">53 skills in this domain</p>

3. code-tour skill is incomplete compared to behuman

behuman ships with a full package: SKILL.md, evals.json, references/, plugin.json, docs page, and mkdocs nav entry.

code-tour is missing:

  • docs/skills/engineering/code-tour.md — no MkDocs documentation page
  • mkdocs.yml nav entry — not discoverable in the docs site
  • .claude-plugin/plugin.json — no plugin manifest (can't be installed via claude /plugin install)
  • evals.json — no test cases to validate trigger behavior

🟢 Minor

4. PR checklist left unchecked

All checklist items were left as [ ]. These should be checked before merging to confirm the contributor verified each item.

5. plugin.json author object includes non-schema field

CLAUDE.md defines the plugin.json schema as allowing only: name, description, version, author, homepage, repository, license, skills. The author object contains a url field:

"author": {
  "name": "Alireza Rezvani",
  "url": "https://alirezarezvani.com"  // ← not in spec
}

Consider flattening to "author": "Alireza Rezvani" to stay within the schema.

6. Content duplication between SKILL.md and docs page

engineering/behuman/SKILL.md and docs/skills/engineering/behuman.md are nearly identical. This creates a maintenance burden — updates must be made in two places. Consider whether the docs page should reference or summarize the SKILL.md rather than duplicating it, or if the docs generation should be scripted (as other skills appear to do via scripts/).


What Works Well ✅

  • behuman/evals.json — 8 test cases covering both positive triggers and negative exclusions. Exactly the right approach for validating skill activation heuristics.
  • Token cost table — clearly communicates the trade-off of running the 3-step loop vs. quiet mode.
  • Anti-patterns table in both skills — specific and actionable, not generic.
  • code-tour SMIG formula — useful structured approach (Situation, Mechanism, Implication, Gotcha) for step descriptions.
  • code-tour validation checklist — the "every file/line must be verified" requirement prevents hallucinated tours.
  • Symlinks and codex index correctly updated for both skills.
  • behuman philosophy section — grounding in Kahneman's dual-process theory is a nice touch and gives the skill conceptual depth.

Suggested Follow-Up

  • Add license: "MIT" to both SKILL.md files
  • Fix engineering skill count to 53
  • Create docs/skills/engineering/code-tour.md and add to mkdocs.yml
  • Add engineering/code-tour/.claude-plugin/plugin.json
  • Add engineering/code-tour/evals.json

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants