MCP HubMCP Hub
Retour aux compétences

install-almanac-content

pjt222
Mis à jour 2 days ago
6 vues
17
2
17
Voir sur GitHub
Métaai

À propos

Cet outil CLI installe des compétences, agents et équipes préconstruits depuis le dépôt agent-almanac dans votre framework agentique pris en charge. Il détecte automatiquement votre framework, résout les dépendances et peut synchroniser les installations à l'aide d'un fichier manifeste. Utilisez-le pour démarrer rapidement des capacités agentiques dans un nouveau projet ou maintenir une configuration cohérente sur plusieurs frameworks.

Installation rapide

Claude Code

Recommandé
Principal
npx skills add pjt222/agent-almanac -a claude-code
Commande PluginAlternatif
/plugin add https://github.com/pjt222/agent-almanac
Git CloneAlternatif
git clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/install-almanac-content

Copiez et collez cette commande dans Claude Code pour installer cette compétence

Documentation

Install Almanac Content

Use agent-almanac CLI to install skills, agents, teams into any supported agentic framework.

When Use

  • Setting up new project, need to install agentic skills, agents, teams
  • Installing all skills from specific domain (r-packages, devops)
  • Targeting multiple frameworks simultaneously (Claude Code, Cursor, Copilot)
  • Creating or syncing declarative agent-almanac.yml manifest for reproducible setups
  • Auditing installed content for broken symlinks or stale references

Inputs

  • Required: Content to install -- one or more skill, agent, team IDs (create-skill, r-developer, r-package-review)
  • Optional: --domain <domain> -- install all skills from domain instead of naming individual IDs
  • Optional: --framework <id> -- target specific framework (default: auto-detect all)
  • Optional: --with-deps -- also install agent skills + team agents+skills
  • Optional: --dry-run -- preview changes without writing to disk
  • Optional: --global -- install to global scope instead of project scope
  • Optional: --force -- overwrite existing content
  • Optional: --source <path> -- explicit path to agent-almanac root (default: auto-detect)

Steps

Step 1: Detect Frameworks

Run framework detection to see which agentic tools present in current project:

agent-almanac detect

Scans working directory for configuration files + directories (.claude/, .cursor/, .github/copilot-instructions/, .agents/) + reports which frameworks active.

Got: Output lists one or more detected frameworks with adapter status. No frameworks detected? Universal adapter (.agents/skills/) used as fallback.

If fail: CLI not found? Ensure installed + on PATH. Detection returns nothing but you know framework is present? Use --framework <id> to specify explicit. Run agent-almanac list --domains to verify CLI can reach registries.

Step 2: Search for Content

Find skills, agents, teams by keyword:

agent-almanac search <keyword>

Browse by category:

agent-almanac list --domains          # List all domains with skill counts
agent-almanac list -d r-packages      # List skills in a specific domain
agent-almanac list --agents           # List all agents
agent-almanac list --teams            # List all teams

Got: Search results or filtered lists display matching content with IDs + descriptions.

If fail: No results appear? Try broader keywords. Verify almanac root reachable: agent-almanac list should show full skill count. Can't find root? Pass --source /path/to/agent-almanac.

Step 3: Install Content

Install one or more items by name:

# Install specific skills
agent-almanac install create-skill write-testthat-tests

# Install all skills from a domain
agent-almanac install --domain devops

# Install an agent with its skills
agent-almanac install --agent r-developer --with-deps

# Install a team with its agents and their skills
agent-almanac install --team r-package-review --with-deps

# Target a specific framework
agent-almanac install create-skill --framework cursor

# Preview without writing
agent-almanac install --domain esoteric --dry-run

# Install to global scope
agent-almanac install create-skill --global

CLI resolves content from registries, selects appropriate adapter for each detected framework, writes files to framework-specific paths (.claude/skills/ for Claude Code, .cursor/rules/ for Cursor).

Got: Output confirms number of items installed + target framework(s). Installed content appears in correct framework directory.

If fail: Items not found? Verify ID matches name field in registry (skills/_registry.yml, agents/_registry.yml, teams/_registry.yml). Files already exist + installation skipped? Use --force to overwrite.

Step 4: Verify Installation

Run health check on all installed content:

agent-almanac audit

Audit specific framework or scope:

agent-almanac audit --framework claude-code
agent-almanac audit --global

See what currently installed:

agent-almanac list --installed

Got: Audit reports all installed items healthy with no broken references. --installed listing shows each item with type + framework.

If fail: Audit reports broken items? Reinstall with --force. Symlinks broken? Verify almanac source path hasn't moved. Run agent-almanac install <broken-id> --force to repair.

Step 5: Manage with Manifest (Optional)

Reproducible setups → use declarative agent-almanac.yml manifest:

# Generate a starter manifest
agent-almanac init

Creates agent-almanac.yml in current directory with detected frameworks + placeholder content lists. Edit file to declare desired skills, agents, teams:

source: /path/to/agent-almanac
frameworks:
  - claude-code
  - cursor
skills:
  - create-skill
  - domain:r-packages
agents:
  - r-developer
teams:
  - r-package-review

Then install everything declared in manifest:

agent-almanac install

Reconcile installed state with manifest (install missing, remove extra):

agent-almanac sync
agent-almanac sync --dry-run  # Preview first

Got: Running install with no arguments reads manifest + installs all declared content. Running sync brings installed state into alignment with manifest, adding missing items + removing undeclared.

If fail: sync reports "No agent-almanac.yml found"? Run agent-almanac init first. Manifest resolves to 0 items? Check skill/agent/team IDs match registry entries exact. Comment lines starting with # ignored.

Step 6: Manage Teams as Campfires (Optional)

Campfire commands provide warm, team-oriented alternative to install --team:

# Browse all available team circles
agent-almanac campfire --all

# Inspect a specific circle (members, practices, pattern)
agent-almanac campfire tending

# See shared agents between teams (hearth-keepers)
agent-almanac campfire --map

# Gather a team (install with arrival ceremony)
agent-almanac gather tending
agent-almanac gather tending --ceremonial    # Show each skill arriving
agent-almanac gather tending --only mystic,gardener  # Partial gathering

# Check fire health (burning / embers / cold)
agent-almanac tend

# Scatter a team (uninstall with farewell)
agent-almanac scatter tending

Campfire state tracked in .agent-almanac/state.json (git-ignored, local to project). Fires have thermal states: burning (used within 7 days), embers (within 30 days), cold (30+ days). Running tend warms all fires + reports health.

Shared skills protected during scatter — if skill needed by another gathered fire, stays installed. Shared agents walk between fires rather than being duplicated.

All campfire commands support --quiet (standard reporter output) + --json (machine-parseable) for scripting.

Got: Teams gathered + managed with state tracking. campfire --all shows fire states. tend reports health.

If fail: Campfire state corrupted? Delete .agent-almanac/state.json + re-gather teams. gather fails? Check team name matches entry in teams/_registry.yml.

Checks

  • agent-almanac detect shows expected frameworks
  • agent-almanac list --installed shows all intended content
  • agent-almanac audit reports no broken items
  • Installed skills resolve in target framework (/skill-name works in Claude Code)
  • Using manifest → agent-almanac sync --dry-run reports no changes needed

Pitfalls

  • Forgetting --with-deps for agents + teams: Installing agent without --with-deps installs only agent definition, not referenced skills. Agent present but unable to follow skill procedures. Always use --with-deps for agents + teams unless already installed dependencies separately.
  • Manifest drift: After manual install or remove content, manifest falls out of sync with actual installed state. Run agent-almanac sync periodically, or always install through manifest to keep aligned.
  • Scope confusion (project vs global): Content installed with --global goes to ~/.claude/skills/ (or equivalent); project-scope content goes to .claude/skills/ in current directory. Skill not found? Check whether installed in wrong scope.
  • Stale source path: agent-almanac repository moved or renamed? --source path in manifests + auto-detection breaks. Update source field in agent-almanac.yml or re-run agent-almanac init.
  • Framework not detected: Detector looks for specific files + directories. Freshly initialized project may not have these yet. Use --framework <id> explicit until project has expected structure, or rely on universal adapter.
  • Campfire thermal state confusion: Fires go cold after 30 days without use. Running agent-almanac tend resets timer for all gathered fires. Fire shows as "cold" → still fully installed — thermal state reflects recency of use, not installation health.

See Also

  • create-skill -- author new skills to add to almanac before installing
  • configure-mcp-server -- set up MCP servers that agents may need after installation
  • write-claude-md -- configure CLAUDE.md to reference installed skills
  • audit-discovery-symlinks -- diagnose symlink issues for Claude Code skill discovery
  • design-cli-output -- terminal output patterns used by CLI reporter + campfire ceremony

Dépôt GitHub

pjt222/agent-almanac
Chemin: i18n/caveman/skills/install-almanac-content
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

Compétences associées

content-collections

Méta

Cette 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.

Voir la compétence

polymarket

Méta

Cette 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.

Voir la compétence

creating-opencode-plugins

Méta

Cette 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.

Voir la compétence

sglang

Méta

SGLang 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.

Voir la compétence