skill-library
- Repo stars 236
- Author updated Live
- Author repo armory
- Domain
- AI
- Compatible agents
-
- Claude Code
- Cursor
- Cline
- Codex
- Windsurf
- Gemini CLI
- +20
- Trust score
- 88 / 100 · community maintained
- Author / version / license
- @Mathews-Tom · no license declared
- Token usage
- Lean
- Setup complexity
- Guided setup
- External API key
- Not required
- Operating systems
- Unspecified (assume cross-platform)
- Runtime requirements
- No special requirements
- Permissions
-
- Read-only
- Write / modify
- Shell exec
- Network behavior
- External requests
- Install commands
- 26 variants
Profile is derived at build time from SKILL.md and install vectors. Subject to drift from author intent.
Heads up: 未限定 allowed-tools，默认拥有全部工具权限。
---
name: skill-library
description: Agent-native catalog and installer for armory packages across all 7 types. Browse, search, insta…
category: ai
runtime: no special runtime
---
# skill-library output preview
## PART A: Task fit
- Use case: Agent-native catalog and installer for armory packages across all 7 types. Browse, search, install, update, and remove without leaving session. Triggers on: "list available packages", "install skill", "armory install", "armory search", "/library list", "package catalog"..
- Inputs: target material, constraints, expected output, and acceptance criteria.
- Evidence boundary: follow “Variables / Supported Package Types / Command Reference” and do not present inference as author intent.
## PART B: Execution result
- **01** The card summarizes the use case; runtime output centers on “Agent-native catalog and installer for armory packages across all 7 types. Browse, search, install, update, and remove without leaving session. Triggers on: "list available packages", "install skill", "armory install", "armory search", "/library list", "package catalog".”.
- **02** When the source has headings, the agent prioritizes “Variables / Supported Package Types / Command Reference” so the result follows the author’s structure.
- **03** Typical output includes task judgment, concrete steps, required commands or file edits, validation, and follow-up options.
- **04** Risk context follows the fingerprint: read files, write/modify files, run shell commands; may access external network resources; usually needs no extra API key.
## Running Rules
- read files, write/modify files, run shell commands; may access external network resources; usually needs no extra API key.
- Validate with a small sample before expanding scope.
- Return the result, validation criteria, and next iteration options.The source mentions slash commands such as `/tmp`, `/library`; use them first when your agent supports command triggers.
Name target files or source material, expected output, forbidden changes, and whether network or shell access is allowed. Permission fingerprint: read files, write/modify files, run shell commands.
Start with a small task and check whether the result follows “Variables / Supported Package Types / Command Reference”. Inspect diffs, logs, previews, or tests before expanding scope.
Confirm the final output includes a concrete result, evidence, and next action. If it stays generic, tighten inputs, boundaries, and acceptance criteria.
---
name: skill-library
description: Agent-native catalog and installer for armory packages across all 7 types. Browse, search, insta…
category: ai
source: Mathews-Tom/armory
---
# skill-library
## When to use
- Agent-native catalog and installer for armory packages across all 7 types. Browse, search, install, update, and remove…
- Use it when the task has clear inputs, repeatable steps, and validation criteria.
## What to provide
- Target material, scope, expected result, and forbidden changes.
- Whether network, commands, file writes, or external services are allowed.
## Execution rules
- Organize steps around “Variables / Supported Package Types / Command Reference” and keep inference separate from source facts.
- read files, write/modify files, run shell commands; may access external network resources; usually needs no extra API key.
- Validate with a small sample before expanding the task.
## Output requirements
- Return the deliverable, key evidence, validation method, and next action.
- Mark missing information as unknown; do not invent commands, platforms, or dependencies.The author source anchors workflow facts; repository files anchor sources and commands; Fluxly only adds fit, limitations, and quality judgment.
skill "skill-library" {
input -> user goal + target files + boundaries + acceptance criteria
context -> Variables / Supported Package Types / Command Reference
rules -> SKILL.md triggers / order / output contract
runtime -> no special runtime | read files, write/modify files, run shell commands | may access external network resources
guardrails -> usually needs no extra API key + small-sample validation + diff/log review
output -> copyable result + checklist + next iteration
}Package Library
Agent-native catalog and installer for all armory package types. Provides browsing, searching, installing, updating, syncing, and removing packages directly within an agent session.
Variables
- ARMORY_REPO:
Mathews-Tom/armory - ARMORY_BRANCH:
main - ARMORY_CATALOG_URL:
https://raw.githubusercontent.com/{ARMORY_REPO}/{ARMORY_BRANCH}/manifest.yaml - DEFAULT_INSTALL_DIR:
~/.claude/ - CATALOG_CACHE_PATH:
/tmp/armory-manifest.yaml - CATALOG_CACHE_TTL:
600(seconds)
Supported Package Types
| Type | Install Target | Method |
|---|---|---|
| skill | ~/.claude/skills/ |
Copy directory |
| agent | ~/.claude/agents/ |
Copy directory |
| hook | ~/.claude/hooks/ |
Copy directory |
| rule | ~/.claude/rules/ |
Body-only file |
| command | ~/.claude/commands/ |
Body-only file |
| utility | ~/.claude/utilities/ |
Copy + chmod +x |
| preset | ~/.claude/presets/ |
Copy directory |
Command Reference
| Command | Cookbook | Purpose |
|---|---|---|
/library list |
cookbook/list.md |
Show all packages with type, version, installed status, update available |
/library list --type <type> |
cookbook/list.md |
Filter listing by package type (skill, agent, hook, rule, etc.) |
/library use <name> |
cookbook/use.md |
Pull a package from armory (auto-detects type from manifest) |
/library search <keyword> |
cookbook/search.md |
Keyword search across all package types, names, and descriptions |
/library search --category <name> |
cookbook/search.md |
Filter search by category (development, review, security, etc.) |
/library sync |
cookbook/sync.md |
Re-pull all installed packages that have updates |
/library info <name> |
cookbook/info.md |
Show full detail for a package (type, version, tags, category) |
/library update |
cookbook/update.md |
Check all installed packages for version bumps (dry-run sync) |
/library remove <name> |
cookbook/remove.md |
Remove an installed package |
/library profiles |
cookbook/profiles.md |
Show available install profiles with package counts |
Cookbook Dispatch
User commands are routed to the corresponding cookbook file based on the subcommand. When a /library command is received, extract the subcommand (the first token after /library) and load the matching cookbook file from the cookbook/ directory relative to this skill. The cookbook file contains the full execution procedure for that operation.
For example, /library use commit-standards dispatches to cookbook/use.md with commit-standards as the package name argument. /library list --type agent dispatches to cookbook/list.md with --type agent as the filter.
If the subcommand does not match any known cookbook, report the error and list the valid subcommands from the table above.
Type Detection
When /library use <name> is called, determine the package type by searching all sections of the manifest (packages.skills, packages.agents, packages.hooks, etc.). The first section containing a matching name entry determines the type. Use the type to:
- Resolve the correct source path:
{type_dir}/{name}/ - Determine the install target:
~/.claude/{install_subdir}/ - Choose the install method: directory copy or body-only extraction
Behavioral Notes
- All bash operations must use absolute paths. Never
cdinto temp dirs. - Fetch uses a 3-tier fallback chain: sparse checkout, then
gh api, thencurl. Attempt each in order; proceed to the next only on failure. - Cache lives in
/tmp/with a 10-minute TTL. A missing cache file is treated as expired (triggers a re-fetch), not as an error. - The catalog is read from
manifest.yamlon GitHub (atARMORY_CATALOG_URL), not from a separate catalog file. This is the same manifest format used by the armory repository. - All package type sections share the same entry format:
name,version,description,path,source, plus optionaltags,category,difficulty.
Decide Fit First
Design Intent
How To Use It
Boundaries And Review