# Audit and Repair Plans Before Execution

Separate plans can hide missing requirements, incompatible interfaces, and broken dependencies. This skill reviews the complete plan set and coordinates repairs with independent verification.

## Install

```bash
npx skillstore add atri10/executor-plan-regression
```

## Metadata

- Status: approved
- Slug: atri10-executor-plan-regression
- Skillstore revision: r1
- Version status: missing
- Tree hash: cc6ed4b97169be03d24a8781ffea8e8526084dfd587d79fc52904843b0d7e709
- Author: atri10
- GitHub username: atri10
- License: MIT
- Repository: https://github.com/atri10/executor/tree/39ddcfe1d9f3497102622b72aa235fb0770187fe/skills/executor-plan-regression
- Ref: 6d0b11444384184b7ae743742a7e233a9a705cd9
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, filesystem
- Quality score: 77
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/atri10-executor-plan-regression
- Manifest: https://skillstore.pages.dev/api/skills/atri10-executor-plan-regression/manifest

## Capabilities

- Builds requirement coverage tables across the complete plan set and identifies missing or invalid coverage claims.
- Checks assumptions and consumed artifacts against producers, interface contracts, and execution order.
- Reviews interface literals, global constraints, shared file writes, and vocabulary across plans.
- Uses plan-lint output and verifies that cited architecture and specification documents exist.
- Directs scoped plan repairs with evidence logs and routes contract amendments to the controller.
- Defines independent re-audits, finding closure tables, human waiver records, and execution-gate verification.

## Use Cases

- Check Plans Before Implementation: Review a drafted plan set for missing requirements and unresolved producer dependencies before developers begin task execution.
- Coordinate Contract Repairs: Assign cross-plan fixes, record specification or interface amendments, and obtain independent confirmation without allowing repairers to grade themselves.
- Verify Readiness Evidence: Compare latest audit reports, repair logs, and human waiver records before confirming that the initiative can enter execution.

## Prompt Templates

### Check Audit Readiness

```
Use executor-plan-regression to check whether my drafted plan set is ready for auditing. Identify missing specifications, contracts, companion tools, and required inputs.
```

### Audit the Complete Plan Set

```
Audit each drafted plan against the complete plan set, specification, and cited interface contracts. Use the supplied auditor template and report all nine checks.
```

### Repair Assigned Findings

```
Coordinate repairs for the failed audit findings. Assign one owner per finding, keep repairs within document boundaries, and dispatch independent re-auditors with repair diffs.
```

### Recheck Changed Plans Before Execution

```
Re-enter plan regression for the changed plans and their affected dependencies. Verify coverage, contract amendments, latest audit references, and human waivers before clearing execution.
```

## Limitations

- Requires companion Executor tools that are referenced but not included in these four files.
- Reviews and repairs planning documents, not implementation code or deployed systems.
- Requires file access, command execution, Git, and separate agent dispatches for the complete workflow.
- Caps audit rounds at three and requires human decisions for unresolved findings or waivers.

## Best Practices

- Provide every plan, its specification, cited contracts, and current lint output before dispatching auditors.
- Keep repairers and re-auditors independent, and preserve each completed audit round.
- Review companion tools before running them, inspect repair diffs, and record human waivers explicitly.

## Anti Patterns

- Auditing one plan without the other plans that supply or consume its outputs.
- Removing requirements or weakening exact interface values to make an audit pass.
- Marking a plan clean from a repair log without a fresh independent audit.

## Security Audit

- Audited at: 2026-10-05T17:55:25.144\+00:00
- Summary: All 243 static findings are false positives involving Markdown backticks, fixed companion-tool paths, or document-review prose. No evidence found of prompt injection, credential exfiltration, or malicious host reconnaissance in the four reviewed files. Referenced companion scripts are absent from this skill, so their implementation and runtime behavior were not verified.

## Stats

- Views: 1
- Downloads: 0
- Favorites: 0
- Popularity score: 0