agentleFS
tikalktikalk/adlc-team-skills/skills/team/team-skills/SKILL.md

team-skills

Skill145 starsChanged 22 days agoMITView on GitHub

Use when listing, adding, or onboarding team skills from the directives repo to the agent's skills directory; --all installs every default and external skill at once.

The file

Show its rules and text

We split this skill into the rules it gives an agent

18 rules. About 0 are common, 5 are distinctive, and 13 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 13

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When the user asks to browse, list, or install team skills.

    line 21
  2. Distinctive:

    When the input names a specific skill to install.

    line 22
  3. Distinctive:

    When extending the agent's capabilities with team-shared skills.

    line 23
  4. Very specific:

    When team-setup delegates /team-skills --all after initial configuration.

    line 24
  5. Very specific:

    Read the file .adlc/init-options.json directly.

    line 34
  6. Very specific:

    If that read fails (file not found), walk up parent directories by reading ../.adlc/init-options.json, then ../../.adlc/init-options.json, …

    line 39
  7. Very specific:

    If team_ai_directives is not set, STOP:

    line 51
  8. Very specific:

    Read {REPO_ROOT}/.skills.json (where {REPO_ROOT} denotes the project root — the directory where .adlc/ lives) and …

    line 65
  9. Distinctive:

    No input: Ask the user which skills to install.

    line 109
  10. Very specific:

    Read .adlc/integration.json to get the current agent

    line 112
  11. Very specific:

    Use {agent_folder}/skills/ as the destination

    line 113
  12. Very specific:

    Check the name against blocked first — refuse immediately if listed.

    line 124
  13. Very specific:

    Verify the entry's url starts with https:// — refuse any other scheme (http://, file://, bare paths) and report it.

    line 125
  14. Distinctive:

    Never execute downloaded content — install means copy the file only.

    line 131
  15. Very specific:

    Never install any skill listed in blocked.

    line 133
  16. Very specific:

    Default skills are installed via --all, typically invoked by team-setup after initial configuration or run manually at any time.

    line 146
  17. Very specific:

    No skill from the blocked section was installed.

    line 177
  18. Very specific:

    Default fallback: team-ai-directives/ relative to project root.

    line 185

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

team-skills

Scope rule: scripts stay skill-local — never source a script from another skill. For another skill's operations, invoke that skill (e.g. team-setup, team-levelup, architect-clarify) via the Skill tool; for its documents, follow the Requires line. scripts/ and references/ below are relative to this skill. {REPO_ROOT} denotes the project root (where .adlc/ lives).

Overview

Browse available skills from the team AI directives and install selected ones to the current agent's skills directory. Use --all to install every default and external skill in one pass. Skills are installed under their original names (no prefix) so they stay in sync with the manifest.

When to Use

  • When the user asks to browse, list, or install team skills.
  • When the input names a specific skill to install.
  • When extending the agent's capabilities with team-shared skills.
  • When team-setup delegates /team-skills --all after initial configuration.
$ARGUMENTS

Core Process

Setup

Read the file .adlc/init-options.json directly. Do NOT use glob, find, or any file-search tool to locate it — search tools may silently skip dotfile-prefixed path segments. Read the file at the exact relative path .adlc/init-options.json from the current working directory.

If that read fails (file not found), walk up parent directories by reading ../.adlc/init-options.json, then ../../.adlc/init-options.json, and so on — up to 4 levels. Stop at the first successful read.

From the JSON, extract the team_ai_directives field:

{
  "team_ai_directives": "/path/to/team-ai-directives"
}

If team_ai_directives is not set, STOP:

Team AI directives not configured.
Run: /team-setup

Path resolution is skill-local:

  • bash scripts/bash/setup-team.sh --json
  • pwsh scripts/powershell/setup-team.ps1 -Json

Step 1: Read Skills Manifest

Read {REPO_ROOT}/.skills.json (where {REPO_ROOT} denotes the project root — the directory where .adlc/ lives) and {TEAM_AI_DIRECTIVES}/.skills.json (team directives manifest). Merge both, with project-root entries taking precedence. Parse the manifest sections (schema v2.0.0):

SectionDescriptionInstalled by --all?
defaultLocal skills (in skills/{name}/) auto-installed during setupYes
externalRemote skills fetched by URL, installed on demandYes
blockedRejected — must never be installedNever

Step 2: Show Available Skills

Present skills grouped by section. For each skill show:

  • Name (the key, e.g. github-actions for local, react-best-practices for external)
  • Description
  • Whether it's already installed (check agent skills directory)

Format:

Team Skills from {TEAM_AI_DIRECTIVES}

Default (local, installed by --all):
  [installed] dbt-template - DBT project templates and best practices
  [available] github-actions - GitHub Actions CI/CD pipeline patterns
  [available] helm-charts - Helm chart patterns for Kubernetes
  [available] crossplane - Crossplane Composition and XRD patterns
  [available] external-secrets - External Secrets Operator patterns
  [available] gke-workload-identity - GKE Workload Identity patterns

External (remote, installed by --all):
  [available] react-best-practices - React performance optimization
  [available] web-design-guidelines - Web interface best practices

Step 3: Install Selected Skills

Determine the install target from the arguments:

  • --all: Install every default skill and every external skill, skipping anything in blocked and anything already installed.
  • A skill name: Install that one skill directly.
  • No input: Ask the user which skills to install.

Determine the agent's skills directory:

  1. Read .adlc/integration.json to get the current agent
  2. Use {agent_folder}/skills/ as the destination

Install under the original name (no prefix), frontmatter name unchanged:

For local skills (default list — names map to skills/{name}/):

  1. Locate {TEAM_AI_DIRECTIVES}/skills/{name}/SKILL.md
  2. Copy SKILL.md to {skills_dir}/{name}/SKILL.md

For external skills (external map — entries have a url):

  1. Check the name against blocked first — refuse immediately if listed.
  2. Verify the entry's url starts with https:// — refuse any other scheme (http://, file://, bare paths) and report it.
  3. Show the user the URL and destination, and ask for confirmation before downloading (preview first, no silent network fetch).
  4. Download SKILL.md from the entry's url field.
  5. Save to {skills_dir}/{name}/SKILL.md.
  6. Never execute downloaded content — install means copy the file only.

Never install any skill listed in blocked. Skip skills already present at the destination (report them as [installed]).

Step 4: Confirm

Installed: github-actions
Location: .opencode/skills/github-actions/SKILL.md
Source: default (local:./skills/github-actions)

Notes

  • Default skills are installed via --all, typically invoked by team-setup after initial configuration or run manually at any time.
  • Skills are installed under their original names — no team- prefix — so the on-disk name matches the manifest key.
  • Skills from the blocked list are never installed.
  • External skills are fetched over HTTPS from the URL in the manifest.

Common Rationalizations

RationalizationWhy it's wrongDo this instead
"I'll add a team- prefix to installed skills"Diverges the on-disk name from the manifest key, breaking lookupsInstall under the original name; frontmatter name stays unchanged
"Search tools will find .adlc/init-options.json"Glob/find silently skip dotfile-prefixed path segmentsRead the exact relative path directly
"Blocked skills are just strong recommendations"blocked means explicitly rejected on installNever install blocked skills
"Default skills already auto-installed elsewhere"Only --all (via team-setup or manual) installs themRun /team-skills --all to onboard the default set
"I'll skip merging the project-root manifest"Project entries are meant to override team defaultsMerge both manifests; project-root wins conflicts

Red Flags

  • Using glob, find, or grep to locate .adlc/init-options.json instead of reading the exact path.
  • Renaming a skill or adding a team- prefix on install — install under the original name so it matches the manifest.
  • Installing a skill listed under the blocked section.
  • Writing skills to any directory other than the agent's configured skills folder.
  • Proceeding past an unset team_ai_directives field instead of stopping with the config message.
  • Fetching an external skill over a non-HTTPS URL.

Verification

  • .adlc/init-options.json was read directly (no search tool) and team_ai_directives resolved to a path.
  • Both {REPO_ROOT}/.skills.json and {TEAM_AI_DIRECTIVES}/.skills.json were merged; project-root entries won any conflicts.
  • Each installed skill lives at {skills_dir}/{name}/SKILL.md with its original name in frontmatter (no prefix).
  • No skill from the blocked section was installed.
  • --all installed every default and external skill (skipping blocked and already-installed).
  • A confirmation block was emitted showing the installed name, on-disk location, and source.

Configuration

  • TEAM_AI_DIRECTIVES — Path to the team AI directives (overrides .adlc/init-options.json).
  • .adlc/init-options.json — Project-level config file with team_ai_directives field.
  • Default fallback: team-ai-directives/ relative to project root.
  • scripts/bash/setup-team.sh / scripts/powershell/setup-team.ps1 — Skill-local path resolution (--json / -Json).

12-Factor Alignment

Factor XII (Team Capability) — manages team skill catalog and installation.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.