skill-composer
Chains multi-step workflows constitutionally — research → plan → judge → execute → verify → log → vault — without human step-by-step prompting
Skill by ariffazil
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/ariffazil-skill-composer
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: research
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/skill-composer
What's inside
- Skill name
skill-composer- Description
- Chains multi-step workflows constitutionally — research → plan → judge → execute → verify → log → vault — without human step-by-step prompting
- License
- AGPL-3.0
- Instructions
- 226 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# skill-composer
**P0 — Autonomous Multi-Step Workflow Chains**
**Seal:** DITEMPA BUKAN DIBERI · 2026-03-27
---
## Purpose
The gap between B+ (81.8%) and A (90%+) is **autonomy** — the ability to chain a multi-step workflow without asking Arif for permission at every step. The `skill-composer` is the orchestration layer that:
1. Takes a sovereign goal (Arif's intent) and decomposes it into constitutional steps
2. Routes each step through the correct skill (research, plan, execute, verify, vault)
3. Enforces F1–F13 at each transition
4. Provides Arif with veto points at the correct moments (F13)
5. Logs everything and seals outcomes to vault
**Why this is P0:** Without this, agents are slaves to the step-by-step prompt. Arif must babysit every research, every code change, every verification. The opencode catastrophe happened because an unanchored agent was given too much rope and no governance layer to stop it from hanging itself.
---
## Workflow Composition Types
### Type A — Research Chain
```
GOAL → deep-research → evidence-map → judgment → report
```
**Used for:** Understanding a system, investigating an issue, fact-finding
**Constitutional gates:** F2 (Truth ≥0.99), F7 (Humility band), F3 (Tri-Witness)
### Type B — Build Chain
```
GOAL → architect-explainer → plan → constitutional-check → forge → postcheck-verifier → vault-writer
```
**Used for:** New features, configuration changes, deployments
**Constitutional gates:** F1 (Reversibility), F11 (CommandAuth), F13 (Sovereign veto)
### Type C — Recovery Chain
```
GOAL → health-probe → diagnosis → rollback-governor → verify → seal
```
**Used for:** Incident response, service recovery, rollback
**Constitutional gates:** F1 (Amanah reversibility), F12 (Injection defense)
### Type D — Audit Chain
```
GOAL → arifos-status → constitutional-check → judgment → report → vault
```
**Used for:** Security audits, constitutional compliance reviews
**Constitutional gates:** F3 (Tri-Witness), F8 (GENIUS gate), F12 (Injection)
---
## Composition Protocol
### Step 0 — Anchor (Always First)
Run `arif-init-wrapper` before any composition.
### Step 1 — Goal Decomposition
For the given goal, identify:
- What skills are needed (in order)
- What constitutional floors are activated at each step
- Where the veto points are (F13 triggers)
- Whether the chain is reversible at each stage
```bash
# Example decomposition for "Fix the broken main branch"
cat << 'EOF'
CHAIN: Recovery Chain (Type C)
STEPS:
1. health-probe → F1, F2 (verify current state)
2. diagnosis → F2, F7 (what is actually broken)
3. architect-explainer → F13 (present plan to Arif)
4. constitutional-check → F1-F13 (full floor review)
5. [VETO POINT] → F13 Arif decides: proceed or not
6. recovery-execute → F1 (reversibility path prepared)
7. postcheck-verifier → F2 (prove outcome, not just logs)
8. vault-writer → F1 (seal the scar)
REVERSIBILITY: Each step has rollback path except steps 6-8.
VETO POINTS: Step 5 (before execution), step 6 (before commit).
EOF
```
### Step 2 — Sequential Execution with Gates
```bash
# Execute each step with constitutional gate
for step in $STEPS; do
# Check constitutional fitness before running
constitutional-check tool=$step 2>/dev/null || {
echo "[GATE FAIL] $step failed constitutional check"
echo "[ACTION] Report to Arif. Do not proceed."
break
}
# Execute step
$step || {
echo "[STEP FAIL] $step failed with exit code $?"
echo "[ACTION] Trigger recovery chain or halt"
break
}
# Verify step outcome
postcheck-verifier step=$step || {
echo "[VERIFY FAIL] $step completed but outcome not proven"
break
}
done
```
### Step 3 — Constitutional Check Before Each Major Transition
Before moving from research → plan, plan → execute, execute → verify:
```bash
# Constitutional gate between phases
arifosmcp_call tool=constitutional_check payload='{From ariffazil/arifOS, AGPL-3.0.
More from ariffazil/arifOS
- 0.669score
memory-archivist
ariffazil
Cross-reference and synthesise daily memory files into patterns, insights, and carry-forward items.
SkillAFreshness: A★ 54Claude Code +9 - 0.669score
arifos-deploy
ariffazil
arifOS sovereign deployment: static hub, docs, runtime, and machine files.
SkillAFreshness: A★ 54Claude Code +9 - 0.664score
arif-a2a-call
ariffazil
A2A v1.0.0 client + 888 JUDGMENT integration for OPENCLAW.
SkillAFreshness: A★ 54Claude Code +9 - 0.664score
openclaw-ops
ariffazil
Complete OpenClaw operations skill: configure models/providers/channels/tools/agents/cron, manage sessions and sub-agents, control paired nodes…
SkillAFreshness: A★ 54Claude Code +9 - 0.661score
postcheck-verifier
ariffazil
Proves actions worked in reality — not just exit code 0, not just logs — by validating actual observable outcomes in the system
SkillAFreshness: A★ 54Claude Code +9 - 0.661score
arifOS-Langfuse
ariffazil
Instrument arifOS constitutional AI kernel with Langfuse LLM tracing.
SkillAFreshness: A★ 54Claude Code +9 - 0.661score
architect-explainer
ariffazil
Translates code, system state, and technical decisions into plain-language decisions, risks, and next actions for Muhammad Arif bin Fazil (non-coder sovereign…
SkillAFreshness: A★ 54Claude Code +9 - 0.661score
AAA Agentic Governance (AAA-Cockpit, canonical)
ariffazil
Governed intelligence skill for AAA as the abstraction, attestation, and abduction control plane across arifOS, APEX, A-FORGE, GEOX, WEALTH, WELL, and the…
SkillAFreshness: A★ 54Claude Code +9 - 0.661score
arif-init-wrapper
ariffazil
Universal session auto-anchor for ALL agent vendors on arifOS VPS — Codex, Claude, Gemini, Kimi, OpenCode, Aider, Copilot, and future agents
SkillAFreshness: A★ 54Claude Code +9 - 0.658score
agi-autonomous-controller
ariffazil
AGI-level autonomous controller — self-healing, self-optimizing, constitutionally governed by arifOS F1-F13
SkillAFreshness: A★ 54Claude Code +9 - 0.658score
arifos-mcp-call
ariffazil
Invoke arifOS constitutional MCP tools (000-999 pipeline, F1-F13 enforced)
SkillAFreshness: A★ 54Claude Code +9 - 0.658score
agentic-governance
ariffazil
Hardened constitutional governance — F1-F13 floor enforcement with Quad-Witness consensus for all actions
SkillAFreshness: A★ 54Claude Code +9
Signals
- Stars
- 54
- Forks
- 9
- Open issues
- 3
- Contributors
- 98
- Last commit (this folder)
- May 11, 2026 (5 months ago)
- Last release
- Oct 17, 2026 (11 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- AGPL-3.0
- Repository
- github.com/ariffazil/arifOS
- Website
- arifos.arif-fazil.com
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from ariffazil
- 0.669score
arifos-deploy
ariffazil
arifOS sovereign deployment: static hub, docs, runtime, and machine files.
SkillAFreshness: A★ 54Claude Code +9 - 0.669score
memory-archivist
ariffazil
Cross-reference and synthesise daily memory files into patterns, insights, and carry-forward items.
SkillAFreshness: A★ 54Claude Code +9 - 0.664score
arif-a2a-call
ariffazil
A2A v1.0.0 client + 888 JUDGMENT integration for OPENCLAW.
SkillAFreshness: A★ 54Claude Code +9 - 0.664score
openclaw-ops
ariffazil
Complete OpenClaw operations skill: configure models/providers/channels/tools/agents/cron, manage sessions and sub-agents, control paired nodes…
SkillAFreshness: A★ 54Claude Code +9 - 0.661score
AAA Agentic Governance (AAA-Cockpit, canonical)
ariffazil
Governed intelligence skill for AAA as the abstraction, attestation, and abduction control plane across arifOS, APEX, A-FORGE, GEOX, WEALTH, WELL, and the…
SkillAFreshness: A★ 54Claude Code +9 - 0.661score
architect-explainer
ariffazil
Translates code, system state, and technical decisions into plain-language decisions, risks, and next actions for Muhammad Arif bin Fazil (non-coder sovereign…
SkillAFreshness: A★ 54Claude Code +9
Cite this page
skill-composer. Agent Mart. https://agentmart.dev/skills/ariffazil-skill-composer (last updated 2026-10-24)