Plan Elixir/Phoenix Feature
Plan a feature by spawning Elixir specialist agents, then output
structured plan with checkboxes.
What Makes /phx-plan Different from /plan
- Spawns Elixir specialist agents for research
- Plans with
[ecto], [liveview], [oban] task routing
- Checks for Iron Law compliance in the plan
- Includes
mix compile/format/credo/test verification
- Understands Phoenix context boundaries
Usage
/phx-plan Add user avatars with S3 upload
/phx-plan .claude/plans/notifications/reviews/notifications-review.md
/phx-plan Implement notifications --depth deep
/phx-plan .claude/plans/auth/plan.md --existing
Arguments
$ARGUMENTS = Feature description, review file, or existing plan
--depth quick|standard|deep = Planning depth (auto-detected)
--existing = Enhance an existing plan with deeper research
Workflow
- Gather context — File path (skip to agents), brainstorm
interview.md (skip clarification), clear description, or vague
- Clarify if vague — Ask questions ONE at a time (skip if
brainstorm interview.md exists with Status: COMPLETE)
- Detect depth — Auto-detect quick/standard/deep
- Runtime context (Tidewave) — Gather live schemas, routes,
and warnings before spawning agents (direct path only — the
research orchestrator gathers its own)
- Spawn research — Selective, based on need. 0–2 agents:
spawn directly in parallel. 3+ agents (broad multi-context
feature): spawn ONE
planning-orchestrator to run and compress
the fan-out, then read only its digest and
summaries/consolidated.md. Create a Claude Code task per spawn:
TaskCreate({subject: "{Agent} research", activeForm: "Researching..."}),
mark in_progress on spawn, completed when done
- Wait for ALL agents — Do NOT proceed until all return
"completed". NEVER write plan while any agent is still running
- Breadboard (LiveView) — System map for multi-page features
- Completeness check — MANDATORY when planning from review
- Split decision — One plan or multiple, concrete options
- Generate plan — Checkboxes, phased tasks, code patterns.
Also create
plans/{slug}/scratchpad.md for decisions and dead-ends
- Self-check (deep only) — Three questions in Risks section
- Present and ask — STOP, show summary, let user decide
When planning from review: Every finding must appear in the
plan — either as a task OR explicitly deferred by the user.
See references/planning-workflow.md for detailed step-by-step.
--existing Mode (Deepening)
Enhances an existing plan instead of creating a new one:
- Load plan, search
.claude/solutions/ for known risks
- Spawn SPECIALIST agents (not Explore) for thin sections.
Each agent writes to
.claude/plans/{slug}/research/ and
returns only a 500-word summary. Same agent selection rules
- Wait for ALL agents (mark tasks
completed as each finishes)
- Add implementation detail, resolve spikes, add verification
- Present diff summary — NEVER delete existing tasks
Iron Laws
- NEVER auto-start /phx-work — Always present plan and ask
- Research before assuming — Web-search unfamiliar tech
- Spawn agents selectively — Only relevant, not all
- NEVER write plan while agents still running
- NEVER skip input findings — Every finding MUST have a task
- Do NOT spawn hex-library-researcher for existing deps
- Skip research when planning from review/investigation — When
input is a review file or
/phx-investigate output, the findings
ARE the research. Do NOT spawn agents to re-discover what the
review already found. Convert findings directly to plan tasks.
(Confirmed: 56-session analysis showed same findings discovered
3-4x across review→investigate→plan phases, wasting ~96K tokens)
Integration with Workflow
/phx-plan {feature} <-- YOU ARE HERE
|
/phx-plan --existing (optional enhancement)
|
ASK USER -> /phx-work .claude/plans/{feature}/plan.md
|
/phx-review → /phx-compound
Notes
- Plans saved to
.claude/plans/{slug}/plan.md
- Research reports in
.claude/plans/{slug}/research/ can be deleted after
CRITICAL: After Writing the Plan
STOP. Do NOT proceed to implementation.
After writing .claude/plans/{slug}/plan.md:
- Summarize: task count, phases, key decisions
- Use
AskUserQuestion with options:
- "Start in fresh session" (recommended for 5+ tasks)
- "Get a briefing" (
/phx-brief — interactive walkthrough)
- "Start here"
- "Review or adjust the plan"
- Wait for user response. Never auto-start work.
When user selects "Start in fresh session", print:
1. Run `/new` to start a fresh session
2. Then run one of:
/phx-work .claude/plans/{slug}/plan.md
/phx-full .claude/plans/{slug}/plan.md (includes review + compound)
This is Iron Law #1. Violating it wastes user context.
References (DO NOT read — for human reference only)
references/planning-workflow.md — Detailed step-by-step
references/plan-template.md
references/complexity-detail.md
references/example-plan.md
references/agent-selection.md
references/breadboarding.md
1---2name: phx-plan-63description: 'Plan features spanning multiple domains: billing (Stripe), auth (RBAC),4 real-time (Presence), webhooks, jobs (Oban). Use when designing interconnected systems5 or converting review findings into tasks.'6metadata:7 source: https://skillmd.com/skills/majiayu000/phx-plan-68---910# Plan Elixir/Phoenix Feature1112Plan a feature by spawning Elixir specialist agents, then output13structured plan with checkboxes.1415## What Makes /phx-plan Different from /plan16171. Spawns Elixir specialist agents for research182. Plans with `[ecto]`, `[liveview]`, `[oban]` task routing193. Checks for Iron Law compliance in the plan204. Includes `mix compile/format/credo/test` verification215. Understands Phoenix context boundaries2223## Usage2425```26/phx-plan Add user avatars with S3 upload27/phx-plan .claude/plans/notifications/reviews/notifications-review.md28/phx-plan Implement notifications --depth deep29/phx-plan .claude/plans/auth/plan.md --existing30```3132## Arguments3334- `$ARGUMENTS` = Feature description, review file, or existing plan35- `--depth quick|standard|deep` = Planning depth (auto-detected)36- `--existing` = Enhance an existing plan with deeper research3738## Workflow39401. **Gather context** — File path (skip to agents), brainstorm41 interview.md (skip clarification), clear description, or vague422. **Clarify if vague** — Ask questions ONE at a time (skip if43 brainstorm interview.md exists with Status: COMPLETE)443. **Detect depth** — Auto-detect quick/standard/deep454. **Runtime context** (Tidewave) — Gather live schemas, routes,46 and warnings before spawning agents (direct path only — the47 research orchestrator gathers its own)485. **Spawn research** — Selective, based on need. **0–2 agents**:49 spawn directly in parallel. **3+ agents** (broad multi-context50 feature): spawn ONE `planning-orchestrator` to run and compress51 the fan-out, then read only its digest and52 `summaries/consolidated.md`. Create a Claude Code task per spawn:53 `TaskCreate({subject: "{Agent} research", activeForm: "Researching..."})`,54 mark `in_progress` on spawn, `completed` when done556. **Wait for ALL agents** — Do NOT proceed until all return56 "completed". NEVER write plan while any agent is still running577. **Breadboard** (LiveView) — System map for multi-page features588. **Completeness check** — MANDATORY when planning from review599. **Split decision** — One plan or multiple, concrete options6010. **Generate plan** — Checkboxes, phased tasks, code patterns.61 Also create `plans/{slug}/scratchpad.md` for decisions and dead-ends6211. **Self-check** (deep only) — Three questions in Risks section6312. **Present and ask** — STOP, show summary, let user decide6465**When planning from review**: Every finding must appear in the66plan — either as a task OR explicitly deferred by the user.6768See `references/planning-workflow.md` for detailed step-by-step.6970### --existing Mode (Deepening)7172Enhances an existing plan instead of creating a new one:73741. Load plan, search `.claude/solutions/` for known risks752. Spawn SPECIALIST agents (not Explore) for thin sections.76 Each agent writes to `.claude/plans/{slug}/research/` and77 returns only a 500-word summary. Same agent selection rules783. Wait for ALL agents (mark tasks `completed` as each finishes)794. Add implementation detail, resolve spikes, add verification805. Present diff summary — **NEVER delete existing tasks**8182## Iron Laws83841. **NEVER auto-start /phx-work** — Always present plan and ask852. **Research before assuming** — Web-search unfamiliar tech863. **Spawn agents selectively** — Only relevant, not all874. **NEVER write plan while agents still running**885. **NEVER skip input findings** — Every finding MUST have a task896. **Do NOT spawn hex-library-researcher for existing deps**907. **Skip research when planning from review/investigation** — When91 input is a review file or `/phx-investigate` output, the findings92 ARE the research. Do NOT spawn agents to re-discover what the93 review already found. Convert findings directly to plan tasks.94 (Confirmed: 56-session analysis showed same findings discovered95 3-4x across review→investigate→plan phases, wasting ~96K tokens)9697## Integration with Workflow9899```text100/phx-plan {feature} <-- YOU ARE HERE101 |102 /phx-plan --existing (optional enhancement)103 |104 ASK USER -> /phx-work .claude/plans/{feature}/plan.md105 |106/phx-review → /phx-compound107```108109## Notes110111- Plans saved to `.claude/plans/{slug}/plan.md`112- Research reports in `.claude/plans/{slug}/research/` can be deleted after113114## CRITICAL: After Writing the Plan115116**STOP. Do NOT proceed to implementation.**117118After writing `.claude/plans/{slug}/plan.md`:1191201. Summarize: task count, phases, key decisions1212. Use `AskUserQuestion` with options:122 - "Start in fresh session" (recommended for 5+ tasks)123 - "Get a briefing" (`/phx-brief` — interactive walkthrough)124 - "Start here"125 - "Review or adjust the plan"1263. Wait for user response. Never auto-start work.127128**When user selects "Start in fresh session"**, print:129130```1311. Run `/new` to start a fresh session1322. Then run one of:133 /phx-work .claude/plans/{slug}/plan.md134 /phx-full .claude/plans/{slug}/plan.md (includes review + compound)135```136137This is Iron Law #1. Violating it wastes user context.138139## References (DO NOT read — for human reference only)140141- `references/planning-workflow.md` — Detailed step-by-step142- `references/plan-template.md`143- `references/complexity-detail.md`144- `references/example-plan.md`145- `references/agent-selection.md`146- `references/breadboarding.md`