try-works/role-model / .agents/skills/recursive-mode
recursive-mode
UnverifiedRepository workflow orchestration skill for staged implementation, locked artifacts, late-phase receipts, and durable memory maintenance. Use when executing recursive-mode runs, resuming a run, locking a phase, or verifying locks.
.agent/PLANS.md
View on GitHub# PLANS.md
<!-- RECURSIVE-MODE-PLANS-BRIDGE:START -->
## recursive-mode plans bridge
This file exists only for tools that expect the Codex plans bridge at `/.agent/PLANS.md`.
The canonical workflow specification lives in `/.recursive/RECURSIVE.md`.
Do not maintain a second authoritative workflow here.
If this bridge conflicts with `/.recursive/RECURSIVE.md`, follow `/.recursive/RECURSIVE.md`.
Short user commands that should trigger recursive-mode orchestration include:
- `Implement the run`
- `Implement run <run-id>`
- `Implement requirement '<run-id>'`
- `Implement the plan`
- `Create a new run based on the plan`
- `Start a recursive run`
Resolution rule:
- If a run id is explicit, use that run.
- If exactly one active/incomplete run exists and no run id is given, resume it.
- If the user refers to a plan, create a new run only when a unique source plan/requirements artifact can be identified from repo docs or immediate task context.
- If the command is ambiguous, ask for the run id or the repo path of the source plan/requirements artifact.
Spec-authoring rule:
- If the user asks to create a plan, help plan, create a spec, or write requirements for a new recursive run, prefer `recursive-spec` before orchestration.
- `recursive-spec` should confirm the user wants spec help, ask what they want to do, read `STATE.md`, `DECISIONS.md`, `MEMORY.md`, and relevant code/tests, keep the draft in temporary non-repo storage, then create the new run only after the requirements are approved.
Benchmark rule:
- If the user asks to benchmark recursive-mode, compare recursive vs non-recursive execution, or generate a recursive-mode benchmark report, install and use the separate optional `recursive-benchmark` add-on on demand instead of assuming benchmark fixtures ship with the default recursive-mode package.
- Prefer `find-skills` when available. Otherwise use `npx skills add <recursive-benchmark-package-or-repo> --full-depth`.
Audit delegation rule:
- If subagents are available and the audit/review context bundle is complete, delegated audit/review is the default path.
- If the controller still chooses `self-audit`, record a concrete `Delegation Override Reason` in the audited phase artifact.
Router rule:
- If the user asks to route delegated work through another transport/model, configure or inspect `/.recursive/config/recursive-router.json`, refresh `/.recursive/config/recursive-router-discovered.json`, re-read both immediately before choosing the delegated CLI/model, and use `recursive-router` before dispatching the delegated role.
<!-- RECURSIVE-MODE-PLANS-BRIDGE:END -->
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |