agentleFS
afittestideafittestide/asimi/.claude/skills/ritual/SKILL.md

ritual

Skill16 starsChanged 7 days agoMITView on GitHub

Author a new Asimi ritual or edit an existing one in .agents/rituals.yaml. Use when the user says "create a ritual", "add a ritual", "new ritual", "edit ritual", or similar.

3 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 3

    1. all-source-os/all-source★ 4same name
    2. haliphax-ai/skills★ 0same name
    3. ritual-foundation/ritual-dapp-skills★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository5 other files it gives its agents.

The file

Show its rules and text

We split this skill into the rules it gives an agent

17 rules. 0 are common, 3 are distinctive, and 14 are very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 14

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before writing YAML, clarify with the user:

    line 12
  2. Very specific:

    If $ARGUMENTS already provides enough detail, skip the questions and proceed.

    line 25
  3. Very specific:

    Add the ritual to .agents/rituals.yaml (create the file if it doesn't exist).

    line 29
  4. Very specific:

    Every ritual needs name, description, and steps

    line 32
  5. Very specific:

    Use inputs for parameters the ritual accepts

    line 33
  6. Very specific:

    Use background to pre-load context shared across multiple steps

    line 34
  7. Very specific:

    Use ritual-level then for outcomes the whole ritual produces

    line 35
  8. Very specific:

    Use act for the prompt — write clear, specific instructions

    line 39
  9. Very specific:

    Use given to pass context variables (e.g.

    line 40
  10. Very specific:

    Use "!command" in given to inject shell command output (e.g.

    line 41
  11. Very specific:

    Use then to declare what a step produces

    line 42
  12. Very specific:

    Use on_failure: retry, on_failure: goto (with on_failure_target), or on_failure: zhengming

    line 44
  13. Very specific:

    Keep act prompts focused — one clear task per step

    line 69
  14. Very specific:

    Use max_retries at the ritual level (default: 1, typical: 2-3)

    line 70
  15. Distinctive:

    After writing the ritual:

    line 75
  16. Distinctive:

    Verify the YAML is valid

    line 76
  17. Very specific:

    Check that all on_failure_target references point to real step names

    line 77

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “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

You are authoring an Asimi ritual. The user's request is: $ARGUMENTS

Gather Requirements

Before writing YAML, clarify with the user:

  1. Goal — What should this ritual accomplish?
  2. Steps — What ministers are involved and in what order? Available ministers:
    • chancellor — coordinates, reports status, requests guidance
    • strategist — analyzes requirements, produces plans
    • forge — implements changes, writes code
    • judge — runs tests, verifies correctness
    • sage — reviews code quality, records precedents
  3. Inputs — Does it need parameters (edict_id, file path, etc.)?
  4. Trigger — Should it activate on an event (e.g. court_started) or be manual (no trigger)?
  5. Failure handling — Should failing steps retry, goto an earlier step, or request zhengming?

If $ARGUMENTS already provides enough detail, skip the questions and proceed.

Write the Ritual

Add the ritual to .agents/rituals.yaml (create the file if it doesn't exist). Follow these rules:

Structure

  • Every ritual needs name, description, and steps
  • Use inputs for parameters the ritual accepts
  • Use background to pre-load context shared across multiple steps
  • Use ritual-level then for outcomes the whole ritual produces

Steps

  • Each step needs a unique name and a minister
  • Use act for the prompt — write clear, specific instructions
  • Use given to pass context variables (e.g. the manifests, the verdicts, the earth status)
  • Use "!command" in given to inject shell command output (e.g. "!just test")
  • Use then to declare what a step produces
  • Steps execute sequentially in array order
  • Use on_failure: retry, on_failure: goto (with on_failure_target), or on_failure: zhengming

Context Variables

Available variables for given and templates:

  • the edict / the edict details — current edict
  • the manifests — file changes
  • the borderlands — unstaged changes
  • the middle kingdom — staged changes
  • the capital — committed changes
  • the verdicts — test results
  • the earth status — full git status (expands to .earth_borderlands, .earth_middle_kingdom, .earth_capital)
  • manifests for the borderlands — diffs for unstaged changes
  • the court status — minister status
  • the unsealed edicts — edicts awaiting approval
  • a heaven's snapshot — heaven's state
  • the project metadata — language, structure info

Patterns

  • Forge/Judge/Sage loop — forge implements, judge tests, sage reviews. On failure, loop back to forging.
  • Strategist first — add a planning step before forging for medium-complexity work.
  • Fork/Join — for parallel work over a list of items. Use fork: { over: step-name, batch_size: N } with a work: block containing sub-steps. Access current item with {{ .item }}.
  • System steps — omit minister for runtime-handled checks (e.g. sandbox health, version checks).

Style

  • Step names should be gerunds (forging, judging, censoring) or descriptive nouns
  • Keep act prompts focused — one clear task per step
  • Use max_retries at the ritual level (default: 1, typical: 2-3)
  • Reference template variables with {{ .variable_name }}

Validate

After writing the ritual:

  1. Verify the YAML is valid
  2. Check that all on_failure_target references point to real step names
  3. Show the user the complete ritual and ask for confirmation before finalizing

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.