skills-registry
Skill3 starsChanged 2 months agoMITView on GitHub
Voltron Lions skills registry and actions map. Use during Skills Gap analysis (Step 4 of the mission flow) to map mission capabilities to registered action types, responsible Lions, and required tools/skills. Also use when the user asks "what can the Lions do", "list voltron actions", or "show the actions map".
In this repository4 other files it gives its agents.
- browser-actionsSkill
- mission-briefSkill
- mission-planSkill
- mission-reportSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 18
Read actions-map.json (use the Read tool on the path relative to the plugin root, or …
- Very specific: line 24
When skill-creator scaffolds a new skill mid-mission, append a new entry to the actions map in your Mission Summary so the user can commit …
- Very specific: line 42
When the user asks directly, render the actions map as a table grouped by Lion: Lion → Action types → What it needs.
≈ “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.
Files
Voltron Lions — Skills Registry
The registry is the single source of truth for which Lion handles which action type and what tools or skills that action requires. It is backed by the machine-readable actions map co-located with this skill:
skills/skills-registry/actions-map.jsonHow to use this registry (Voltron Main — Step 4)
During Skills Gap analysis:
- Read
actions-map.json(use theReadtool on the path relative to the plugin root, or${CLAUDE_PLUGIN_ROOT}/skills/skills-registry/actions-map.jsonwhen the plugin root variable is available). - For each capability the mission needs, find the closest matching
actionentry:- Exact or close match found,
status: "covered"→ classify the capability ascovered, cite the responsible Lion from the entry. - Match found with non-empty
skillslist → classify asuse-existing-skill <name>using the first skill in the list; the responsible Lion invokes it. - Match found with
requiresfield → the capability is conditionally covered. Surface the requirement (e.g., per-site user approval forlive-browser-action) in the Skills Gap table Notes column. - No match → classify as
needs-new-skilland propose a one-line description.
- Exact or close match found,
- When
skill-creatorscaffolds a new skill mid-mission, append a new entry to the actions map in your Mission Summary so the user can commit it (the plugin's copy ofactions-map.jsonis read-only at runtime — registry updates ship with the next plugin version).
Actions map schema
Each entry in actionTypes:
| Field | Meaning |
action | Kebab-case action type identifier (e.g., api-integration) |
lion | The responsible Lion agent (red-lion, blue-lion, green-lion, yellow-lion, or voltron-main) |
description | One line: what this action type covers |
requiredTools | Built-in Claude Code tools the Lion needs (must be in the Lion's frontmatter tools:) |
skills | Skills the Lion invokes for this action (empty = built-ins suffice) |
requires | Optional precondition outside this action (e.g., per-site approval, or a prior document) |
status | covered | conditional | experimental |
Answering "what can the Lions do"
When the user asks directly, render the actions map as a table grouped by Lion:
Lion → Action types → What it needs. Keep it under 30 lines; link to actions-map.json for the full data.
Registering a new action type
New action types are added by editing actions-map.json in the plugin repository (https://github.com/Cloudscockpit/voltron-lions-claudecode), bumping the plugin version, and reinstalling. An entry MUST name exactly one responsible Lion — if an action seems to need two Lions, split it into two entries with a dependsOn note in the description.
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
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.