evolve-skill
À propos
La compétence `evolve-skill` met à jour une compétence existante en affinant son contenu actuel ou en créant une nouvelle variante avancée. Elle gère l'intégralité du processus, depuis l'évaluation et la collecte des exigences jusqu'à l'application des modifications, en passant par la mise à jour des métadonnées de version et des références croisées. Utilisez-la lorsque les étapes d'une compétence sont obsolètes, que des retours révèlent des lacunes ou qu'une amélioration de la complexité est nécessaire.
Installation rapide
Claude Code
Recommandénpx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/evolve-skillCopiez et collez cette commande dans Claude Code pour installer cette compétence
Documentation
Evolve an Existing Skill
Fix, grow, or make advanced variant of skill first made with create-skill. This proc covers upkeep side of skill life: check gaps, apply tight fixes, bump versions, keep registry and cross-refs in sync.
When Use
- Skill proc steps stale or thin after tool shifts
- User feedback shows missing pitfalls, unclear steps, or weak check
- Skill need grow from basic to intermediate (or intermediate to advanced)
- Advanced variant needed next to original (e.g.,
create-r-packageandcreate-r-package-advanced) - Related skills added or removed and cross-refs are stale
Inputs
- Required: Path to existing SKILL.md to evolve
- Required: Evolve trigger (feedback, tool change, complexity bump, new related skills, spotted pitfalls)
- Optional: Target complexity if changing (basic, intermediate, advanced)
- Optional: Make advanced variant vs refine in-place (default: refine in-place)
Steps
Step 1: Assess the Current Skill
Read existing SKILL.md and check each section vs quality list:
| Section | What to Check | Common Issues |
|---|---|---|
| Frontmatter | All required fields present, description < 1024 chars | Missing tags, stale version |
| When to Use | 3-5 concrete trigger conditions | Vague or overlapping triggers |
| Inputs | Required vs optional clearly separated | Missing defaults for optional inputs |
| Procedure | Each step has code + Expected + On failure | Missing On failure blocks, pseudocode instead of real commands |
| Validation | Each item is binary pass/fail | Subjective criteria ("code is clean") |
| Common Pitfalls | 3-6 with cause and avoidance | Too generic ("be careful") |
| Related Skills | 2-5 valid skill references | Stale references to renamed/removed skills |
# Read the skill
cat skills/<skill-name>/SKILL.md
# Check frontmatter parses
head -20 skills/<skill-name>/SKILL.md
# Verify related skills still exist
grep -oP '`[\w-]+`' skills/<skill-name>/SKILL.md | sort -u
Got: List of specific gaps, weak spots, or fix chances.
If fail: SKILL.md not exist or no frontmatter? This skill not apply — use create-skill instead to make from scratch.
Step 2: Gather Evolution Requirements
Spot and sort what fired evolve:
| Trigger | Example | Typical Scope |
|---|---|---|
| User feedback | "Step 3 is unclear" | Refinement |
| Tooling change | New API version, deprecated command | Refinement |
| Discovered pitfall | Common failure not documented | Refinement |
| Complexity upgrade | Skill is too shallow for real use | Refinement or variant |
| New related skills | Adjacent skill was added | Refinement (cross-refs) |
| Advanced use case | Power users need deeper coverage | Variant |
Log specific changes needed before edit. List each change with target section.
Got: Concrete list of changes (e.g., "Add On failure to Step 4", "Add new Step 6 for edge case X", "Update Related Skills to include new-skill").
If fail: Changes unclear? Ask user for clarity before go on. Vague evolve goals give vague fixes.
Step 3: Choose Evolution Scope
Use this pick matrix to decide refine in-place or make variant:
| Criteria | Refinement (in-place) | Advanced Variant (new skill) |
|---|---|---|
| Skill ID | Unchanged | New ID: <skill>-advanced |
| File path | Same SKILL.md | New directory |
| Version bump | Patch or minor | Starts at 1.0 |
| Complexity | May increase | Higher than original |
| Registry | No new entry | New entry added |
| Symlinks | No change | New symlinks needed |
| Original skill | Modified directly | Left intact, gains cross-reference |
Refinement: Pick when fixing quality, patching gaps, or adding modest new content. Skill keeps its identity.
Variant: Pick when evolved version would double length, change target audience, or need different inputs. Original stays as-is for simpler use cases.
Got: Clear pick — refine or variant — with reason.
If fail: Unsure? Default to refine. Can always pull variant later; harder to merge one back.
Step 4: Apply Content Changes
For Refinements
Edit existing SKILL.md direct:
# Open for editing
# Add/revise procedure steps
# Strengthen Expected/On failure pairs
# Add tables or examples
# Update When to Use triggers
# Revise Inputs if scope changed
Follow these edit rules:
- Keep all existing sections — add content, do not remove sections
- Keep step numbering sequential after inserts
- Every new or modified step must have both Expected and On failure
- New pitfalls go at end of Common Pitfalls section
- New related skills go at end of Related Skills section
For Variants
# Create the variant directory
mkdir -p skills/<skill-name>-advanced/
# Copy the original as a starting point
cp skills/<skill-name>/SKILL.md skills/<skill-name>-advanced/SKILL.md
# Edit the variant:
# - Change `name` to `<skill-name>-advanced`
# - Update `description` to reflect the advanced scope
# - Raise `complexity` (e.g., intermediate → advanced)
# - Reset `version` to "1.0"
# - Add/expand procedure steps for the advanced use case
# - Reference the original in Related Skills as a prerequisite
Got: SKILL.md (refined or new variant) passes check list from Step 1.
If fail: Step edit breaks doc shape? Use git diff to review changes and revert partial edits with git checkout -- <file>.
Step 4.5: Sync Translated Variants
Required when translations exist. This step applies to both human authors and AI agents following this procedure. Do not skip — stale
source_commitvalues causenpm run validate:translationsto report false staleness warnings across all locales.
Check whether translations exist for evolved skill and update to match new source state:
# Check for existing translations
ls i18n/*/skills/<skill-name>/SKILL.md 2>/dev/null
If translations exist
- Get current source commit hash:
SOURCE_COMMIT=$(git rev-parse HEAD)
- Update
source_commitin each translated file frontmatter:
for locale_file in i18n/*/skills/<skill-name>/SKILL.md; do
sed -i "s/^source_commit: .*/source_commit: $SOURCE_COMMIT/" "$locale_file"
done
- Flag files for re-translation by adding affected locales in commit msg:
evolve(<skill-name>): <description of changes>
Translations flagged for re-sync: de, zh-CN, ja, es
Changed sections: <list sections that changed>
- Regen translation status files:
npm run translation:status
If no translations exist
No action needed. Go to Step 5.
For variants
Wait translation of new variants until variant stabilizes (1-2 versions). Translating v1.0 variant that may change much by v1.2 wastes effort. Add translations after variant refined at least once.
Got: All translated files have source_commit updated to current commit. Commit msg notes which locales need re-translation and which sections changed. npm run translation:status exits 0.
If fail: sed fails to match frontmatter field? Translated file may have odd format. Open by hand and check it has source_commit in its YAML frontmatter. Field missing? File not scaffolded right — re-scaffold with npm run translate:scaffold.
Step 5: Update Version and Metadata
Bump version field in frontmatter by semver:
| Change Type | Version Bump | Example |
|---|---|---|
| Typo fix, wording clarification | Patch: 1.0 → 1.1 | Fixed unclear sentence in Step 3 |
| New step, new pitfall, new table | Minor: 1.0 → 2.0 | Added Step 7 for edge case handling |
| Restructured procedure, changed inputs | Major: 1.0 → 2.0 | Reorganized from 5 to 8 steps |
Also update:
complexityif scope grew (e.g., basic → intermediate)tagsif coverage area shifteddescriptionif skill scope materially different
Got: Frontmatter version shows size of changes. New variants start at "1.0".
If fail: Forget to bump version? Next evolve will have no way to tell current state from old. Always bump before commit.
Step 6: Update Registry and Cross-References
For Refinements
No registry changes needed (path unchanged). Update cross-refs only if Related Skills shifted in other skills:
# Check if any skill references the evolved skill
grep -r "<skill-name>" skills/*/SKILL.md
For Variants
Add new skill to skills/_registry.yml:
- id: <skill-name>-advanced
path: <skill-name>-advanced/SKILL.md
complexity: advanced
language: multi
description: One-line description of the advanced variant
Then:
- Bump
total_skillsat top of registry - Add Related Skills cross-ref in original skill pointing to variant
- Add Related Skills cross-ref in variant pointing to original
- Make symlinks for slash command discovery:
# Project-level
ln -s ../../skills/<skill-name>-advanced .claude/skills/<skill-name>-advanced
# Global
ln -s /mnt/d/dev/p/agent-almanac/skills/<skill-name>-advanced ~/.claude/skills/<skill-name>-advanced
Got: Registry total_skills matches find skills -name SKILL.md | wc -l. Cross-refs two-way.
If fail: Registry count wrong? Run find skills -name SKILL.md | wc -l to get true count and fix registry. Broken symlinks? Use readlink -f to debug.
Step 7: Validate the Evolved Skill
Run full check list:
- SKILL.md exists at expected path
- YAML frontmatter parses with no errors
-
versionwas bumped (refinement) or set to "1.0" (variant) - All sections present: When to Use, Inputs, Procedure, Validation, Common Pitfalls, Related Skills
- Every procedure step has Expected and On failure blocks
- Related Skills ref valid, existing skill names
- Registry entry exists with right path (variants only)
-
total_skillscount matches real skill count on disk - Symlinks resolve right (variants only)
-
git diffshows no slip deletes from original content - For refinements with translations:
source_commitupdated or translations flagged for re-sync
# Verify frontmatter
head -20 skills/<skill-name>/SKILL.md
# Count skills on disk vs registry
find skills -name SKILL.md | wc -l
grep total_skills skills/_registry.yml
# Check symlinks (for variants)
ls -la .claude/skills/<skill-name>-advanced
readlink -f .claude/skills/<skill-name>-advanced/SKILL.md
# Review all changes
git diff
Got: All check items pass. Evolved skill ready to commit.
If fail: Fix each failing item one by one. Most common post-evolve issue is stale total_skills count — always check it last.
Validation
- SKILL.md exists and has valid YAML frontmatter
-
versionfield shows changes made - All procedure steps have Expected and On failure blocks
- Related Skills refs valid (no broken cross-refs)
- Registry
total_skillsmatches real count on disk - For variants: new entry in
_registry.ymlwith right path - For variants: symlinks made at
.claude/skills/and~/.claude/skills/ -
git diffconfirms no slip content removal - For refinements with translations:
source_commitupdated or translations flagged for re-sync
Pitfalls
- Forget to bump version: No version bump, no way to track what changed or when. Always update
versionin frontmatter before commit. - Slip content delete: Restructure steps, easy to drop On failure block or table row. Always review
git diffbefore commit. - Stale cross-refs: Make variant, both original and variant need ref each other. One-way refs leave graph broken.
- Registry count drift: After make variant,
total_skillscount must bump. Forget this cause check fails in other skills that test registry. - Stale translations after evolve: With 1,288 translation files in repo, every skill evolve fires staleness in up to 4 locale files. Always check for existing translations with
ls i18n/*/skills/<skill-name>/SKILL.mdand updatesource_commitin each translated file frontmatter, or flag them for re-translation in commit msg. Skip this causenpm run validate:translationsto report stale warnings. - Scope creep during refine: Refine that doubles skill length should probably be variant instead. Adding more than 3 new procedure steps? Rethink scope pick from Step 3.
- Avoid
git mvon NTFS-mounted paths (WSL): On/mnt/paths,git mvfor directories can make broken perms (d?????????). Usemkdir -p+ copy files +git rmold path instead. See environment guide troubleshoot section.
See Also
create-skill— base for making new skills; evolve-skill assumes this was followed firstcommit-changes— commit evolved skill with clear msgconfigure-git-repository— version-tracked skill changessecurity-audit-codebase— review evolved skills for slip-added secrets
Dépôt GitHub
Compétences associées
content-collections
MétaCette compétence propose une configuration éprouvée en production pour Content Collections, un outil axé sur TypeScript qui transforme des fichiers Markdown/MDX en collections de données typées de manière sûre avec une validation Zod. Utilisez-la lors de la création de blogs, de sites de documentation ou d'applications Vite + React riches en contenu pour garantir la sécurité de typage et la validation automatique du contenu. Elle couvre tout, de la configuration du plugin Vite et de la compilation MDX à l'optimisation des déploiements et la validation des schémas.
polymarket
MétaCette compétence permet aux développeurs de créer des applications avec la plateforme de marchés prédictifs Polymarket, incluant l'intégration d'API pour le trading et les données de marché. Elle fournit également une diffusion de données en temps réel via WebSocket pour surveiller les transactions en direct et l'activité du marché. Utilisez-la pour mettre en œuvre des stratégies de trading ou pour créer des outils traitant les mises à jour de marché en direct.
creating-opencode-plugins
MétaCette compétence aide les développeurs à créer des plugins OpenCode qui s'interconnectent avec plus de 25 types d'événements tels que les commandes, les fichiers et les opérations LSP. Elle fournit la structure du plugin, les spécifications de l'API événementielle et les modèles d'implémentation pour les modules JavaScript/TypeScript. Utilisez-la lorsque vous avez besoin d'intercepter, de surveiller ou d'étendre le cycle de vie de l'assistant IA OpenCode avec une logique personnalisée pilotée par les événements.
sglang
MétaSGLang est un framework de service LLM haute performance spécialisé dans la génération rapide et structurée pour les workflows JSON, regex et agentiques grâce à son cache de préfixe RadixAttention. Il offre une inférence nettement plus rapide, particulièrement pour les tâches avec des préfixes répétés, ce qui le rend idéal pour les sorties complexes et structurées ainsi que les conversations multi-tours. Choisissez SGLang plutôt que des alternatives comme vLLM lorsque vous avez besoin d'un décodage contraint ou que vous construisez des applications avec un partage étendu de préfixes.
