Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/tttg2010/tj-shortnpx agentmods add skills/tttg2010/tj-short/skillWrote 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/tttg2010/tj-short/skill)<a href="https://agentmods.dev/skills/tttg2010/tj-short/skill"><img src="https://agentmods.dev/badge/skills/tttg2010/tj-short/skill/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/tttg2010/tj-short/skill"><img src="https://agentmods.dev/badge/skills/tttg2010/tj-short/skill.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 4 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Privilege Escalation · line 633 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
- high Privilege Escalation · line 644 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
- high Privilege Escalation · line 671 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
- high Privilege Escalation · line 680 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
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 | $0.00112 | $0.11975 |
| Opus 5.5 | $0.00045 | $0.04790 |
| Sonnet 5.5 | $0.00022 | $0.02395 |
| Haiku 4.5 | $0.00011 | $0.01197 |
Grade A, and why
tj-short 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 1mo 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 — 863 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TJ Short
TJ Short is a Codex Skill for ecommerce short dramas.
Version: short-drama-ecommerce v0.8.30
v0.8.30 Changelog
- Added
Script Fidelity and Traceability Ruleafter two real production failures: a shot table silently dropped 2 of 4 script dialogue lines with no shot ever picking them up, and a first-frame generation added an unrequested prop (a physical invitation card the script never described) that survived review and was inherited into the final video. - Added dialogue/action ID tagging (
[D-<scene>-<n>],[A-<scene>-<n>]) required in the episode script, and a coverage ledger required at the end of Phase 2 mapping every ID to a shot or a stated deliberate cut. 100% coverage is a hard gate before Phase 3. - Added a
SOURCE ANCHORblock requirement: every Phase 4/5 prompt document must open with the exact tagged script lines the shot derives from, so generation happens against the source, not a paraphrase of a paraphrase. - Added a containment diff check for every image/video generation: prompts must carry MUST HAVE and MUST NOT HAVE lists, and review must answer a mandatory, non-skippable "what did the output add that the prompt never asked for?" field. First-frame drift is a hard gate (cheap retake); video-level drift is a soft gate (record + human decides) unless it violates an explicit MUST NOT HAVE item, which always escalates to blocking.
- Added
保真核对.mdto Required Deliverables: one file per episode carrying the coverage ledger and containment diffs. - Clarified this adds no dedicated LLM calls: coverage is deterministic ID-string matching, and containment diffs are filled during review that already happens when a human looks at generated output.
- Added
US-Market Localization Hard Rule: when target region is 美国 / US / United States, the story, dialogue, characters, scene text, subtitles, CTA, compliance wording, and project manifest must fit the United States market instead of translating Chinese ecommerce short-drama habits into English. - Added current production split: salpx/gpt-image-2 is the default image layer for storyboard grids and first frames; MiniMax H3 is the default video layer for raw clips. salpx video models remain optional alternatives when the user selects them.
- Source: architecture review by 肖老师 (xiao-laoshi-agent) after two script-fidelity failures surfaced in a real project; see
docs/methodology.mdSection 18 for the full diagnosis and design rationale.
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.
- 1mo ago Changed · +106 lines · +14 tokens per session f168c9c33b43
- 1mo ago First seen · 757 lines · 98 tokens per session scan A bfd536125756
tj-short is a skill published in the GitHub repository tttg2010/tj-short (18 stars, last pushed 1mo ago), licensed MIT. It adds 112 tokens to every session and 11,975 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-08-30.
Other skills, from other repositories
daiso-cli
A command-line interface for searching Daiso and related Korean shopping, store, food, cinema, and convenience-store data through the Daiso project. It can return structured JSON for searches and comparisons.
gc-still-image-motion-director
Analyze still images and design restrained, image-specific motion for image-to-video generation. Use when the user provides one or more images, screenshots, posters, collages, illustrations, editorial graphics, or an image folder and asks how the image should move, which elements must stay fixed, whether it should…
barcode-get-barcode
Returns a raw PNG image of the barcode. Ideal for direct embedding or file download.
douyin-drama-commerce
A Chinese-language workflow for planning product sales through Douyin, a Chinese short-video and livestream platform. It connects product evidence with drama concepts, creator selection, scripts, publishing checks, compliance review, and performance analysis.
ecommerce-detail-page-generator
Turn product photos and a minimal brief into verified, platform-adapted e-commerce detail-page systems, including product fact review, visual analysis, localized copy, 6-10 module story planning, policy-aware text-free image generation or real-photo routing, deterministic Pillow typography, platform-sized slices…
ucp
Use when the user wants to use the UCP CLI to find, compare, buy, or track products from online merchants, or to set up and troubleshoot the local UCP profile required for merchant-scoped operations. Covers global catalog search ("find me X under $Y"), named-merchant transactions ("buy this from Z.com"), order…