pirfl

pirfl is a skill for a fallibilist Plan → Implement → Review → Fix → Loop workflow.

The package keeps the user-requested name pirfl for easy explicit invocation, but the skill’s description also exposes clearer trigger terms: plan-implement-review-fix, error-correcting workflow, validation loop, subagent review, iterative critique, and multi-task orchestration.

Contents

pirfl/
├── SKILL.md
├── references/
│   ├── cf-principles.md
│   ├── loop-protocol.md
│   ├── multi-task-orchestration.md
│   ├── research-notes.md
│   ├── subagent-prompts.md
│   └── task-log-format.md
├── assets/
│   ├── pirfl-log-template.md
│   ├── reviewer-report-template.md
│   ├── standalone-error-correction-loop-skill.template.md
│   └── task-portfolio-template.md
├── scripts/
│   ├── new_pirfl_log.py
│   └── validate_skill.py
└── evals/
    ├── quality-evals.json
    └── trigger-evals.json

Install

Place the pirfl/ folder wherever your skills-compatible agent loads skills, such as a user-level or repo-level skills directory. The folder name and name: field both use pirfl for compatibility.

Validate

python pirfl/scripts/validate_skill.py pirfl

Create a work log

python pirfl/scripts/new_pirfl_log.py "Refactor payment flow" --out pirfl-log.md

Optional companion skill

assets/standalone-error-correction-loop-skill.template.md is a template for a narrower standalone loop skill named error-correction-loop. It is intentionally not named review-and-fix and is not stored as SKILL.md inside this package so the main ZIP remains a single-skill bundle.

License

Dual licensed under the Unlicense and CC0 1.0 Universal. See LICENSE.