A pre-publish QA pass for AI-assisted or human drafts, checking the things a spellchecker won't catch.
- Inputs: draft content, target keywords, brand guidelines, and source materials to fact-check against
- Flags potential AI hallucinations, unsupported claims, and factual inaccuracies
- Runs a citation audit for missing or needed sources
- Checks style consistency and brand voice alignment
- Outputs an originality score with flagged passages plus a prioritized list of revisions
- Run in deep-research mode so claims get checked against the actual source materials
Paste a draft agent skill or reusable prompt along with your project context, the target task, available tools, and known risks, and get back a structured audit covering purpose, required inputs, output format, safety boundaries, and failure modes.
The review also flags missing checks and repeatability gaps, then returns an improved version ready to test — useful before wiring a skill into an autonomous coding agent or handing it to a team.
Submit a proposal along with your project context, stack, constraints, and team's skill level to get a skeptical second opinion before committing engineering weeks to it.
The review separates what's genuinely strong from what's over-engineered, surfaces hidden coupling that will resurface later as maintenance cost, and ends with a clear verdict plus a minimal implementation path instead of just a list of concerns.
Acts as a pre-publish SEO gate for a prompt-library page, scoring
frontmatter and body content before it goes live.
- Inputs: page frontmatter, page content, target keyword, audience
- Checks title clarity, meta description quality, heading structure, and
search intent match
- Flags thin-content risk and missing examples or inputs
- Outputs a 1-10 score, an issues table, fixes, and rewritten frontmatter
Reviews a full draft against its audience, goal, tone target, and publishing channel, then delivers direct coaching-style feedback instead of generic encouragement.
Covers what works, what fails, where readers will get confused, which sentences to cut, a stronger structure recommendation, and a rewritten version of the weakest section. Best used on important pages before they go live.
Paste a full video script along with its audience, goal, target length, and tone, and this prompt audits it the way a YouTube retention editor would — flagging a weak hook, slow or redundant sections, missing context, shaky transitions, unclear claims, and CTA placement. It follows up with prioritized fixes, a rewritten hook, sample rewritten transitions, and a shortened version of the script if the pacing calls for it.
Paste a diff to get a fast, senior-engineer-style pass that prioritizes what actually matters — correctness bugs, regressions, security issues, missing tests, and maintainability risks — over style nitpicks. Each finding comes with severity, evidence pulled directly from the diff, and a concrete suggested change, making it a solid gate to run before opening a pull request.
Feed in a suspicious snippet along with the error message and expected behavior, and get back a ranked list of likely bugs instead of a vague hunch. Each finding names the observable symptom, the probable root cause, the exact lines or functions responsible, and a concrete fix to apply. When the pasted context isn't enough to be confident, it says exactly what extra detail is missing instead of guessing at a diagnosis.