Full Agent Skill

Full Phoenix Feature Development

Execute complete Elixir/Phoenix feature development autonomously: research patterns,

plan with specialist agents, implement with verification, Elixir code review.

Cycles back automatically if review finds issues.

Usage

/phx:full Add user authentication with magic links
/phx:full Real-time notification system with Phoenix PubSub
/phx:full Background job processing for email campaigns --max-cycles 5
/phx:full Add magic-link auth --codex

Wrong input guard: if the argument is a path to an existing plan file

(.claude/plans/*/plan.md), do NOT re-plan it. Say so and run /phx:work {path}

instead — the plan phase already happened.

Flags

• --codex — Codex CLI joins the review panel on every review

cycle, matching /phx:review --codex.

• Requires the codex CLI. A missing CLI degrades to a SKIPPED note —

the cycle never fails.

• Consensus findings (flagged by a Claude agent and codex) are

marked HIGH CONFIDENCE.

Workflow Overview

┌──────────────────────────────────────────────────────────────────┐
│                       /phx:full {feature}                        │
├──────────────────────────────────────────────────────────────────┤
│                                                                  │
│  ┌────────┐  ┌────────┐  ┌────────┐  ┌────────┐  ┌────────┐  ┌────────┐  │
│  │Discover│→ │  Plan  │→ │  Work  │→ │ Verify │→ │ Review │→ │Compound│→Done│
│  │ Assess │  │[Pn-Tm] │  │Execute │  │  Full  │  │4 Agents│  │Capture │     │
│  │ Decide │  │ Phases │  │ Tasks  │  │  Loop  │  │Parallel│  │ Solve  │     │
│  └───┬────┘  └────────┘  └────────┘  └───┬────┘  └────────┘  └────────┘     │
│       │                            ↑      │    ↑              │         │
│       ├── "just do it" ────────────┤      │    │              │         │
│       ├── "plan it" ──┐            │      ↓    │              │         │
│       │               ↓            │ ┌────────┐│              │         │
│       │     ┌──────────────┐       │ │Fix     ││ ┌─────────┐ │         │
│       │     │   PLANNING   │       │ │Issues  │└─│ Fix     │←┘         │
│       │     └──────────────┘       │ └───┬────┘  │ Review  │           │
│       │                            │     ↓       │ Findings│           │
│       │                       ┌────┴─────────┐   └────┬────┘           │
│       │                       │   VERIFYING   │←──────┘                │
│       └── "research it" ─────┘  (re-verify)                            │
│            (comprehensive plan)                                         │
│                                                                  │
│  On Completion:                                                  │
│  Auto-compound: Capture solved problems → .claude/solutions/     │
│  Auto-suggest: /phx:document → /phx:learn-from-fix                       │
│                                                                  │
└──────────────────────────────────────────────────────────────────┘

State Machine

STATES: INITIALIZING → DISCOVERING → PLANNING → WORKING →
        VERIFYING → REVIEWING → COMPLETED → COMPOUNDING | BLOCKED

Claude Code nesting compatibility

Before delegating this cycle, determine the effective maximum nesting depth:

• If CLAUDE_CODE_MAX_SUBAGENT_SPAWN_DEPTH is a positive integer, use it.
• Otherwise, inspect claude --version: the default is 1 in Claude Code

2.1.217–2.1.218 and 3 in 2.1.219+.

• If the version cannot be determined, conservatively use 1.

• Depth 3+ — the full nested topology is available; delegate to

phx:workflow-orchestrator as usual.

• Depth 1–2 — execute the same state machine in this main

session. Read ${CLAUDE_PLUGIN_ROOT}/agents/workflow-orchestrator.md for the

phase contract, but spawn leaf research/review specialists directly. Do not

spawn workflow-orchestrator, planning-orchestrator, or

parallel-reviewer, because those agents need to delegate again. Context

supervision may still be spawned directly after its input files exist.

The fallback is a topology change only: preserve the same user decisions,

artifacts, verification gates, retry/cycle limits, and review-to-fix loop.

Never tell the user to set an environment variable just to make /phx:full

work.

Save state in .claude/plans/{slug}/progress.md AND via Claude Code

tasks. Create one task per phase at start, mark in_progress on

entry and completed on exit:

TaskCreate({subject: "Discover & assess complexity", activeForm: "Discovering..."})
TaskCreate({subject: "Plan feature", activeForm: "Planning..."})
TaskCreate({subject: "Implement tasks", activeForm: "Working..."})
TaskCreate({subject: "Verify implementation", activeForm: "Verifying..."})
TaskCreate({subject: "Review with specialists", activeForm: "Reviewing..."})
TaskCreate({subject: "Capture solutions", activeForm: "Compounding..."})

Set up blockedBy dependencies between phases (sequential).

Run COMPOUNDING phase on COMPLETED to capture solved problems in .claude/solutions/.

Suggest /phx:document for docs and /phx:learn-from-fix for quick pattern capture.

Cycle Limits

| Setting | Default | Description |

|---------|---------|-------------|

| --max-cycles | 10 | Max plan→review cycles |

| --max-retries | 3 | Max retries per task |

| --max-blockers | 5 | Max blockers before stopping |

Stop with INCOMPLETE status when limits exceeded. List remaining work and recommended action.

Integration

/phx:full = /phx:plan → /phx:work → /phx:verify → /phx:review → (fix → /phx:verify) → /phx:compound

Use Ralph Wiggum Loop for fully autonomous execution:

/ralph-loop:ralph-loop "/phx:full {feature}" --completion-promise "DONE" --max-iterations 50

Iron Laws

• NEVER skip verification — Every task must pass mix compile --warnings-as-errors before moving to the next. Run mix test <affected> per-phase, full suite only at final gate
• Respect cycle limits — When --max-cycles is exhausted, STOP with INCOMPLETE status. Do not continue indefinitely hoping the next fix works
• One state transition at a time — Follow the state machine strictly. Never jump from PLANNING to REVIEWING — each state produces artifacts the next state needs
• Discover before deciding — Always run DISCOVERING phase to assess complexity. Skipping it for "simple" features leads to underplanned implementations
• Agent output is findings, not fixes — Review agents report issues. Only the WORKING state makes code changes
• Skip redundant review agents — In REVIEWING phase: skip

verification-runner (work phase already verified), skip iron-law-judge

if PostToolUse hooks verified all files. For <200 lines changed,

spawn only elixir-reviewer + security-analyzer (if auth files)

• ZERO narration in autonomous mode — This is a HARD rule, not

a suggestion. NEVER write "Let me now...", "Now I need to...",

"I'll now...", "Next, I will...", or any preamble before a tool

call. Just call the tool. Only output text for: decisions that

need explanation, errors, or phase transitions. If you catch

yourself narrating, delete the text and just make the tool call.

(Post-PR validation: 30% of messages still violated this — the

instruction was too soft. This stronger wording is required.)

References

• ${CLAUDE_SKILL_DIR}/references/execution-steps.md — Detailed step-by-step execution
• ${CLAUDE_SKILL_DIR}/references/example-run.md — Example full cycle run
• ${CLAUDE_SKILL_DIR}/references/safety-recovery.md — Safety rails, resume, rollback
• ${CLAUDE_SKILL_DIR}/references/cycle-patterns.md — Advanced cycling strategies