Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add crevideo/crevideo-reach --skill performance-diagnosisgit clone --depth 1 https://github.com/crevideo/crevideo-reachWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/crevideo/crevideo-reach/performance-diagnosis)<a href="https://agentmods.dev/skills/crevideo/crevideo-reach/performance-diagnosis"><img src="https://agentmods.dev/badge/skills/crevideo/crevideo-reach/performance-diagnosis/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/crevideo/crevideo-reach/performance-diagnosis"><img src="https://agentmods.dev/badge/skills/crevideo/crevideo-reach/performance-diagnosis.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00128 | $0.01514 |
| Opus 5 | $0.00064 | $0.00757 |
| Sonnet 5 | $0.00026 | $0.00303 |
| Haiku 4.5 | $0.00013 | $0.00151 |
Grade A, and why
performance-diagnosis scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Performance Diagnosis · 效果诊断(洞察引擎)
Role: turn affiliate report data into decisions. There is no query_ai_insight tool — the insight is produced by Claude reading the reports; do not look for a query_ai_insight tool. Read-only, no outbound. Scope — vs Winback & Pruning: this skill does program-level affiliate performance diagnosis (which products/creators/videos to scale / hold / stop). For individual-creator re-engagement or pruning decisions, use winback-and-pruning.
Output language
Write every output in the merchant's working language, using that market's native seller terminology:
- US Local sellers → English: tier / all-in take-rate / outreach waterfall / Target Collaboration / effective creators …
- China POP sellers → 中文: 分级 / 真实抽成 / 触达阶梯 / 定向邀约 / 有效达人 …
Tool names (e.g.
query_collaboration_performance) stay identical in both languages. If unsure which market, ask once before producing output.
When to use / not use
- Use: health-check running automations, review GMV/conversion, data diagnosis for a weekly/monthly digest, "insight + recommendations".
- Don't use: handling specific replies (→ reply-triage); a single automation's failure post-mortem (→ winback-and-pruning).
Inputs
- shop_cipher + time range (days, default 30, or start/end).
- Focus (optional): overall / a product / a video / graduation / win-back.
Steps (in order)
- Fetch (read-only): query_collaboration_performance (scope target/open/both), query_product_performance, query_shoppable_video_performance. Use view=overview/breakdown/detail as needed.
- Field traps: product/video breakdown+detail are TT-official REAL-TIME (match the brand-app page) while overview/trend/collab are legacy aggregates — small overview-vs-breakdown drift is expected (two data generations), note it instead of "reconciling".
*_incrementis the period total (not a growth rate) but applies to the legacy views only — TT-official rows use plain names (gmv/orders/units_sold). TT-official lists are token-paginated (you get the first page + a "Showing N of M" line; narrow the date range for the rest — there is no full-set dump). Responses carry "data as of" (latest_available_date, 1-2 day lag) — a missing "today" is not zero performance. Write N/A for any number you can't get and say which query failed — never fabricate. - Explicit reasoning: before giving conclusions, think through "data → meaning → likely cause" internally (insight is produced by your own reasoning, not by some ready-made insight tool); avoid zero-shot guessing.
- Compute the 5 leading indicators (definitions in the Cheat Sheet): ① qualified acceptance rate (meaningful interaction, not raw reply rate) ② reply → first-video time ③ graduatable creator count ④ effective-creator share (active creators producing converting content) ⑤ outreach-pressure / suppression health (share blocked by cooldown / already-replied rules; rising = drifting toward spam). Give each a current value + a trend arrow.
- Risk scan: all-in take-rate over the line (threshold in the registry), suspected duplicate outreach, zero-output automations (0 accepted / 0 videos), violation signs.
- Decision-style output: each point lands on change → cause → decision needed → recommended action → risk of inaction → evidence; anything that can't become a decision is "too descriptive" — drop it.
- Close with "N items need your decision now".
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 66 lines · 128 tokens per session scan A b8d11b0be701
performance-diagnosis is a skill published in the GitHub repository crevideo/crevideo-reach (7 stars, last pushed 3d ago), licensed MIT. It adds 128 tokens to every session and 1,514 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
markifact-overview
Reference — what Markifact is, what the MCP server exposes, and the discover→inspect→run pattern. Always loaded into the performance-marketer agent.
safe-write-operations
Reference — rules for safely executing write/destructive operations against ad accounts. Always loaded into the performance-marketer agent.
dev
Software development planning. Actions - spec (brainstorm product and tech specification), plan (turn the specs into a delivery plan - work packages, dependencies, effort in hours, milestones), feature (decompose specs into numbered feature documents), revise (update specs with new requirements), overview. Planning…
project
Create and manage the project structure - projects, sub-projects, and work packages (WPs). Actions - new project, new subproject, new WP, overview. Hierarchy depth (2 or 3 levels) is fixed at creation. All work happens in WPs; projects and sub-projects are pure structure.
ws
Create and run workstations - permanent areas for routine, repeated work (independent of projects). Actions - new WS, new subWS, new area, new routine, run, from project, list, overview. Hierarchy depth (2 or 3 levels) is fixed at creation.
cease-desist
Draft a cease-and-desist letter (send mode) or triage one you received (receive mode). Use when asserting your rights against an infringer with a demand letter calibrated to your enforcement posture, or when an incoming C&D needs triage into a structured options memo with a recommendation.