pt-study-sop

Analyze a project's past Codex sessions, memory files, and existing local skills to recommend the highest-value skills to create or update. Use when a user asks what skills a project needs, wants skill ideas grounded in real project history, wants an audit of current project-local skills, or wants recommendations for updating stale or incomplete skills instead of creating duplicates.

Apply treys-swarm-planner to the PT Study System by grounding in repo canon, the active execution board, Conductor tracks, planner APIs, and the `study_tasks` schema. Use when planning significant work in this repo and then executing it end-to-end with a backward-built wave/task graph until the goal is complete, unless the user explicitly asks for plan-only or review-only output.

Implements Manus-style file-based planning for complex tasks. Creates task_plan.md, findings.md, and progress.md. Use when starting complex multi-step tasks, research projects, or any task requiring >5 tool calls.

Execute tasks from a track's implementation plan following TDD workflow

Manage track lifecycle: archive, restore, delete, rename, and cleanup

Create a new track with specification and phased implementation plan

Git-aware undo by logical work unit (track, phase, or task)

Execute dependency-aware tasks from a plan using Codex subagents in parallel. Re-check unblocked tasks after each wave, enforce worker prompt quality, and validate outputs before integrating results.

Build a swarm-ready implementation plan with phases, task dependencies, explicit deliverables, and verification gates. Use when work is multi-step, spans many files, or can benefit from parallel Codex subagents.

Diagnose the user's real execution bottlenecks, contradictions, and leverage points, then build a sequenced operating system around the true constraint. Use when the user wants blunt accountability, blind-spot analysis, a hard-truth diagnosis, a 90-day operating plan, or asks what is actually holding them back.

Produce a hardened implementation plan and require a worker-based second-opinion review before accepting it. Use when stakes are high, scope is large, or drift risk is elevated.

General-purpose X/Twitter research agent. Searches X for real-time perspectives, dev discussions, product feedback, cultural takes, breaking news, and expert opinions. Works like a web research agent but uses X as the source. Use when: (1) user says "x research", "search x for", "search twitter for", "what are people saying about", "what's twitter saying", "check x for", "x search", "/x-research", (2) user is working on something where recent X discourse would provide useful context (new library releases, API changes, product launches, cultural events, industry drama), (3) user wants to find what devs/experts/community thinks about a topic. NOT for: posting tweets, account management, or historical archive searches beyond 7 days.

Git worktree based parallel agent workflow for running Codex CLI, Claude Code, OpenCode, and other coding agents concurrently without file conflicts. Use when you need to (1) create/manage named worktrees (integrate/ui/brain[/docs]), (2) route a task to the correct worktree based on the files it will touch, (3) launch a tool in the chosen worktree, or (4) coordinate multiple agents via an integrator merge and review loop.

Spawn Codex subagents via background shell to offload context-heavy work. Use for: deep research (3+ searches), codebase exploration (8+ files), multi-step workflows, exploratory tasks, long-running operations, documentation generation, or any other task where the intermediate steps will use large numbers of tokens.

Ensure global agent instruction files include the standard '# Workflow' block; update idempotently and report a unified diff.