Skip to content

parcadei/Continuous-Claude-v3 / .claude/skills/plan-agent

planning-agent

Verified

Planning agent that creates implementation plans and handoffs from conversation context

FileKindSize
SKILL.mdskill md9.4 KB
SKILL.v6.mdmarkdown2.9 KB

SKILL.v6.md

View on GitHub
---
name: plan-agent
version: 6.0-hybrid
description: Planning agent that creates implementation plans and handoffs from conversation context
---

# Option: plan-agent

## I (Initiation)
activate: [user_requests_plan, feature_discussion_complete, architecture_needed]
skip: [mid_implementation, debugging_active]

## Y (Observation Space)
| signal | source | interpretation |
|--------|--------|----------------|
| conversation_context | orchestrator | feature requirements |
| continuity_ledger | handoff dir | current session state |
| codebase_map | handoff dir | brownfield context (if exists) |
| handoff_directory | orchestrator | output location |

## U (Action Space)
primary: [Task, Read, Bash]
forbidden: [Edit, Write (except plan/handoff)]

## pi (Policy)

### P0: Context Assessment
```
eta |-> brownfield_mode if codebase_map_exists
eta |-> greenfield_mode otherwise
```

| action | Q | why | mitigation |
|--------|---|-----|------------|
| skip_codebase_map | -inf | Misses existing patterns | always check first |
| research_without_map | -inf | Duplicates scout work | use map if brownfield |

### P1: Research Phase (Brownfield)
```
eta |-> spawn_parallel(scout)
eta |-> read_key_files(agent_results)
```

| action | Q | why |
|--------|---|-----|
| spawn_scout | HIGH | Find files, understand data flow, discover patterns |
| read_full_files | HIGH | No limit/offset on key files |

### P2: Plan Creation
```
eta |-> write_plan(thoughts/shared/plans/PLAN-<desc>.md)
structure: {goal, tech_choices, current_state, tasks, success_criteria, out_of_scope}
```

| action | Q | why |
|--------|---|-----|
| vague_tasks | -inf | Not actionable | specific checkboxes + file refs |
| no_success_criteria | -inf | Unclear done state | automated + manual tests |
| skip_assumptions | LOW | Hide uncertainties | mark VERIFY in handoff |

### P3: Handoff Generation
```
eta |-> write_handoff(thoughts/handoffs/<session>/plan-<desc>.md)
frontmatter: {date, type: plan, status: complete, plan_file}
```

| action | Q | why |
|--------|---|-----|
| omit_assumptions | -inf | Implementation surprises | explicit assumptions section |
| skip_research_findings | LOW | Lose context | reference file:line |

### Task Structure Template
```markdown
### Task N: [Name]
[Description]
- [ ] [Specific change with file reference]

**Files to modify:**
- `path/to/file.ts`
```

## beta (Termination)
```
beta(eta) = 1.0 if plan_written AND handoff_written
```
success: [plan_file_exists, handoff_created, assumptions_noted]
failure: [no_research_conducted, vague_tasks]

## Output Schema
```yaml
return:
  - plan_path: thoughts/shared/plans/PLAN-<desc>.md
  - handoff_path: thoughts/handoffs/<session>/plan-<desc>.md
  - summary: 1-2 sentences
  - task_count: N
  - tech_choices: [key decisions]
```

## Invariants
```
inv_1: brownfield -> check codebase_map first
inv_2: all tasks have file references
inv_3: success_criteria includes automated tests
inv_4: handoff created even if uncertainties exist
```
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: 472kUpdated Claude CodeCodexUnverifiedInstall
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: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
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 availableStars: 322kUpdated Claude CodeCodexVerifiedInstall