autoplan

name	autoplan
description	Auto-review pipeline — runs strategy, engineering, and security review in sequence and produces an implementation plan with auto-decisions. Used at the start of any In Progress issue.
allowed-tools	Bash, Read, Grep, Glob, Skill
user-invocable	true
argument-hint	[issue-number]
automation	manual
metadata	{"version":"1.0","created":"2026-04-28T00:00:00.000Z","author":"Ability.ai","changelog":["1.0: Initial version — runs strategy, engineering, and security review in sequence and produces an implementation plan with auto-decisions for an In Progress issue."]}

Autoplan

ℹ️ First, set expectations: before anything else, print one short line with this skill's version and its most recent change — the top entry of metadata.changelog above — e.g. autoplan vX.Y — recent: <summary>. Then proceed.

Run the full pre-implementation review pipeline for an issue: strategy review, engineering review, and security review. Produces a structured plan with auto-decisions so the developer can approve or revise before implementation begins.

State Dependencies

Source	Location	Read	Write	Description
GitHub Issue	GitHub repo	✅		Issue title, body, acceptance criteria
Requirements	docs/memory/requirements.md	✅		Feature context
Architecture	docs/memory/architecture.md	✅		System design
Feature Flows	docs/memory/feature-flows/	✅		Existing patterns

Arguments

• $ARGUMENTS: GitHub issue number (e.g., 42 or #42). If omitted, reads from current branch name or asks.

Process

Step 1: Load Issue Context

ISSUE=${ARGUMENTS#\#}
gh issue view $ISSUE --json number,title,body,labels,assignees,milestone

Extract:

• Problem statement
• Acceptance criteria
• Priority and type labels
• Milestone or epic association

Step 2: Strategy Review

Assess the issue at a product and architecture level:

1. Scope clarity — Is the acceptance criteria unambiguous? Flag if not.
2. Dependencies — Does this block or depend on other open issues?
3. Risk level — P0/P1 flags, auth/security surface, data migrations
4. Approach options — Enumerate 2-3 approaches; select the simplest that satisfies requirements
5. Auto-decision: Record chosen approach and rationale

Step 3: Engineering Review

Translate the chosen approach into a concrete implementation plan:

1. Files to modify — identify specific files with line ranges
2. Files to create — new files and their purpose
3. Data model changes — schema/migration changes if any
4. API surface changes — new or modified endpoints
5. Test plan — what tests are needed (happy path, auth, edge cases)
6. Effort estimate — Small / Medium / Large
7. Auto-decision: Record implementation order and test strategy

Step 4: Security Review

Identify security considerations before writing code:

1. Auth boundaries — Does this touch authentication or authorization?
2. Input vectors — New endpoints, file uploads, user-controlled data?
3. Data exposure — Can this leak PII, credentials, or internal state?
4. Dependency risk — New packages being introduced?
5. Auto-decision: Flag any items requiring explicit security attention during implementation

Step 5: Produce Plan Document

Output the plan in this format:

## Autoplan: Issue #[N] — [Title]

### Strategy Decision
**Chosen approach**: [description]
**Rationale**: [why]
**Risk level**: [Low / Medium / High]

### Implementation Plan
**Files to modify:**
- `path/to/file.py` — [what changes]

**Files to create:**
- `path/to/new.py` — [purpose]

**Order of changes:**
1. [step]
2. [step]

**Test plan:**
- [ ] [test case]
- [ ] [test case]

### Security Flags
- [flag or "None identified"]

### Auto-Decisions
| Decision | Choice | Confidence |
|----------|--------|------------|
| Approach | [name] | High/Medium/Low |
| Test strategy | [name] | High/Medium/Low |

### Human Review Required
- [ ] Approve approach
- [ ] Approve implementation plan
- [ ] Acknowledge security flags

Completion Checklist

• Issue context loaded
• Strategy review completed with approach selected
• Engineering review completed with implementation plan
• Security review completed with flags identified
• Plan document output
• Human review checklist presented