compose
Skill3 starsChanged 6 days agoMITView on GitHub
Internal compose engine; load only via stage holder HARD-GATE — do not invoke directly.
13 other projects use a version of this.
Adopted by
Most recently taken up by
- tlyboy/skillstheir own version, first seen Nov 1, 2026
- jantichy/claudetheir own version, first seen Oct 27, 2026
- by-carlos/daikenjatheir own version, first seen Oct 13, 2026
- matthewharrilal/claude-skillstheir own version, first seen Oct 13, 2026
- drader/researcher_agenttheir own version, first seen Oct 13, 2026
- flexprice/flexpricetheir own version, first seen Oct 8, 2026
Used with
Projects with it also use review.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
SentrySkillreview
Fixture review skill.
Also in 3 projects that use this
Other versions13 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 13
- flexprice/flexprice★ 6.9ksame name
- Meet-Miyani/compose-skill★ 303same name
- renatoworks/strudel-claude★ 61same name
- jantichy/claude★ 45same name
- Epistates/sparX★ 3same name
- drader/researcher_agent★ 2same name
- cacardinal/laila★ 1same name
- tlyboy/skills★ 1same name
- by-carlos/daikenja★ 0same name
- csfh/skills★ 0same name
3 more not listed
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 repository11 other files it gives its agents.
- decisionSkill
- eval-librarySkill
- lulu-approachSkill
- lulu-betSkill
- lulu-brainstormSkill
- lulu-designSkill
- lulu-execSkill
- lulu-planSkill
- lulu-specSkill
- lulu-tasksSkill
- lulu-workflowSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 2 are distinctive, and 5 are very specific to this project.
0
Common
Most of the 12 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 51
Run $START_COMPOSE. Load {SKILL_ROOT}/compose/references/compose-ontology.md once.
- Very specific: line 55
Run $SESSION_INFO --view session, then bind:
- Distinctive: line 76
Ask: Deliver or Modify.
- Very specific: line 83
Run $SESSION_INFO --view delivery-preview.
- Distinctive: line 84
Wait for explicit delivery confirmation.
- Very specific: line 86
Run $SESSION_CONTROL deliver --confirm.
- Very specific: line 87
Run $SESSION_INFO --view stage-transitions.
“Common” is measured: kept by at least 6 in 10 of the 12 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.
Files
compose
Shared compose engine for stage holders. It takes one input document, runs one
execution, and delivers one document. Step execution lives in
references/execution.md.
Inputs
| Variable | Required | Source | Use |
$CYCLE_ID | yes | Holder runtime foundation | Active cycle |
$PROFILE_PATH | yes | Holder preflight | Runtime compose-profile.json |
$SCOPE_PACKAGE | yes | Holder preflight | Normalized scope-package.json |
Script Macros
Macro expansion: {SKILL_ROOT}/_runtime.md § Script Macros → Macro expansion. Non-zero exit → Blocking: stop, report (stderr / exit code), wait for user direction.
| Macro | Command |
$START_COMPOSE | python3 "$SKILL_ROOT/compose/scripts/session/start.py" --cycle-id "$CYCLE_ID" --profile-path "$PROFILE_PATH" --scope-package "$SCOPE_PACKAGE" |
$SESSION_INFO | python3 "$SKILL_ROOT/compose/scripts/session/session_info.py" --cycle-id "$CYCLE_ID" --view <view> |
$SESSION_CONTROL | python3 "$SKILL_ROOT/compose/scripts/session/session_control.py" --cycle-id "$CYCLE_ID" <subcommand> |
Subcommands and stdout: script module docstrings or --help.
Execution macros are defined in {SKILL_ROOT}/compose/references/execution.md. Load that file before using them.
Lifecycle
stateDiagram-v2
[*] --> Working: Start
Working --> ReadyForDelivery: execution Completed
ReadyForDelivery --> Working: Modify
ReadyForDelivery --> Delivered: DeliverRoute only from $SESSION_INFO stdout. Do not invent the next step.
Entry
Start
Run $START_COMPOSE. Load {SKILL_ROOT}/compose/references/compose-ontology.md once.
Bind context
Run $SESSION_INFO --view session, then bind:
| Cite | JSON field | Use |
$REVISION_DIR | revision_dir | revision-scoped tools |
$DEMAND_MANIFEST | demand_manifest | Delivery demand atomization; skip when null |
$CURRENT_STATE | workflow_state.current_state | Outer routing |
$STEP_STATE | execution.state | Step routing |
Done: $CURRENT_STATE=Working.
Working
Session state: Working.
Load {SKILL_ROOT}/compose/references/execution.md and follow it.
Done: $STEP_STATE=Completed → ## ReadyForDelivery.
ReadyForDelivery
Ask: Deliver or Modify.
- Deliver → ## Delivery.
- Modify →
$SESSION_CONTROL return-to-working --confirm. Then ## Working. Preview is not required.
Delivery
- Run
$SESSION_INFO --view delivery-preview. On failure → Blocking. Show the preview; full compose document only if asked. - Wait for explicit delivery confirmation.
- Demand manifest (only when
$DEMAND_MANIFESTis present): enumerate delivered demands per$DEMAND_MANIFEST.unit_rule, then$SESSION_CONTROL write-demand-manifest --units-json '<JSON array>'. Skip when$DEMAND_MANIFESTis null. On failure → Blocking. - Run
$SESSION_CONTROL deliver --confirm. On failure → Blocking (including leftoveragenda.jsonblockers). - Run
$SESSION_INFO --view stage-transitions. On non-zero exit → Blocking. Prompt next stages when present.
Reference
| Document | When |
{SKILL_ROOT}/compose/references/compose-ontology.md | Start, once per session |
{SKILL_ROOT}/compose/references/execution.md | Working |
{SKILL_ROOT}/eval/SKILL.md | Loaded from execution Evaluating |
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.