fli-review

Reviews project state against a submission deadline by verifying claims against live data.

25|1|Updated Feb 10, 2026
One-click install
npx skills add https://github.com/wisdom-in-a-nutshell/agents --skill fli-review-wisdom-in-a-nutshell
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: fli-review
Source: https://github.com/wisdom-in-a-nutshell/agents/tree/main/skills-source/owned/fli-review
Command: npx skills add https://github.com/wisdom-in-a-nutshell/agents --skill fli-review-wisdom-in-a-nutshell

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Project documentation drifts from reality, and self-review misses blind spots. This Skill acts as an independent senior reviewer that loads project context cold, verifies claims against live data and running systems, and critiques everything against one question: does this work push the case study toward earning the next interview by the submission deadline? ## Core Features & Use Cases - Cold-context loading: Reads the case prompt, status docs, project trackers, architecture overview, and recent git history in a defined order to rebuild current project state. - Claim verification: Runs the repo's own check script, queries live databases, inspects the running product, and spot-checks citations instead of trusting documentation. - Endpoint-anchored critique: Applies five lenses (endpoint drift, demo truth, defensibility, blind spots, sequencing) and delivers ranked risks, next moves, and explicit disagreements. - Use Case: Before a weekly planning session, ask for a review to discover that a claimed data-coverage metric does not match the live database and that current work is off the critical path to the demo. ## Quick Start Use fli-review to load the current project context, verify recent claims against live data, and give me an endpoint-anchored critique with top risks and next moves.

Frequently Asked Questions about fli-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I get an independent review of my project before a deadline?▼

Invoke the review to load context cold from status docs, task trackers, and git history, then receive a critique judged against your submission endpoint. The output includes verified-versus-claimed findings, ranked risks, and ordered next moves.

How to verify project documentation matches actual code and data?▼

The review runs the repo's own check script, queries live SQLite databases directly, inspects the running product, and spot-checks that citations resolve to real evidence. At least one claim per review is verified against ground truth rather than docs.

Can this reviewer write code or fix the issues it finds?▼

No. The role is critic and auditor, not implementer. It identifies risks, blind spots, and disagreements with reasoning, and only records decisions or build-log entries when explicitly asked.

What does the review output look like?▼

The output has five parts: a short state summary, verified versus claimed items, top risks ranked by threat to the deadline, one to three justified next moves including what not to do, and direct disagreements with recorded decisions.

When should I not use an endpoint-anchored project review?▼

Avoid it when you need implementation work, quick code fixes, or a review untethered to a concrete deliverable. It is designed for milestone-driven projects with a defined rubric and deadline, not general-purpose code review.