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 agentmods add instructions/jnhflow21/social-media-toolkit/agents-mdgit clone --depth 1 https://github.com/JNHFlow21/social-media-toolkitWrote 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/instructions/jnhflow21/social-media-toolkit/agents-md)<a href="https://agentmods.dev/instructions/jnhflow21/social-media-toolkit/agents-md"><img src="https://agentmods.dev/badge/instructions/jnhflow21/social-media-toolkit/agents-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.01026 | $0.01026 |
| Opus 5 | $0.00513 | $0.00513 |
| Sonnet 5 | $0.00205 | $0.00205 |
| Haiku 4.5 | $0.00103 | $0.00103 |
Grade C, and why
social-media-toolkit AGENTS.md scanned grade C with 1 finding 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 5d 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.
Harvests environment variableshighData exfiltration
Enumerating or grepping the environment for keys collects credentials unrelated to what the mod says it does.
11. The NPX path must be a real tested installer, not a documentation alias. It must not collect credentials, load project secrets, or duplicate Python runtime behavior in JavaScript. 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.
Agent Development Protocol
This repository is a public, reusable product. It must not depend on the Brain vault or any private user workspace.
Product boundary
- Core package:
social_media_toolkit. - Thin MCP transport:
social_post_extractor_mcp. - Thin npm bootstrap:
package.jsonplusbin/social-media-toolkit.mjs; it may installuvand must delegate the actual CLI/MCP installation touv tool. - Public read paths must not require browser automation, CDP, Playwright, or a logged-in session.
- Account-private analytics remain optional and isolated.
- Publishing/upload automation belongs in a separate future package because it has authenticated side effects.
Invariants
- Plain canonical-text precedence is GetNote original content → native platform subtitle → Volcengine cloud ASR.
- Timed YouTube transcripts are a separate evidence contract: manual subtitle cues → automatic subtitle cues → timestamped Volcengine cloud ASR. Callers may explicitly force Volcengine ASR, optionally with anonymous speaker diarization and bounded public-metadata context; that mode must bypass native captions. Non-timestamped GetNote text must not short-circuit either route.
- Non-empty GetNote original content wins even when a task still contains a stale error message on the plain canonical-text route.
- Douyin media resolution uses the free public adapter first. Only after that adapter fails, and only when
TIKHUB_API_KEYis configured, may the single orchestrator call TikHub for temporary public CDN URLs. Preserve the paid route and ephemeral-URL warning in provenance. inspectand plaintextdo not download persistent media. Timed text requires an explicit output directory and may write only requested transcript artifacts; media remains temporary. Downloads require an explicit output directory.- Every normalized result preserves provenance, warnings, platform limitations, and paid-route disclosure.
- SDK, CLI, and MCP must call the same
SocialMediaToolkit; do not add legacy aliases or a second scheduler. - Volcengine is the only ASR provider. TikHub resolves media only; do not treat it as a text/ASR provider or add local ASR/another cloud fallback.
- Never claim a public comment sample is a global platform ranking.
- A plain
text, text-enabledcapture, timed-text, or full-chain smoke-test request runs its configured text route automatically. Do not ask for a second authorization; report the route and any possible usage charge afterward. - The NPX path must be a real tested installer, not a documentation alias. It must not collect credentials, load project secrets, or duplicate Python runtime behavior in JavaScript.
- The current MCP transport uses the v1
FastMCPAPI and must keepmcp>=1.28.1,<2until a dedicated MCP 2.x migration lands. CI must execute a real stdio initialize handshake, not just compile the module.
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.
- 5d ago First seen · 66 lines · 1,026 tokens per session scan C 79f1f92aec47
social-media-toolkit AGENTS.md is an instructions file published in the GitHub repository JNHFlow21/social-media-toolkit (30 stars, last pushed 4d ago), licensed Apache-2.0. It adds 1,026 tokens to every session, about $0.0051 per session on Opus 5. A static security scan graded it C with 1 finding (harvests environment variables). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
douyin-transcript-mcp AGENTS.md
AGENTS.md instructions for mei-shui-xing/douyin-transcript-mcp, covering instructions for ai coding agents, required workflow and non-negotiable boundaries.
openrouter-mcp-multimodal AGENTS.md
AGENTS.md instructions for stabgan/openrouter-mcp-multimodal, covering agent instructions, before you ship, releasing (read this before publishing), short version and version files (must all match package.json).
llm-context.py CLAUDE.md
Instructions for cyberchitta/llm-context.py, covering claude.md, working notes (gitignored) and draining the field notes.
seekstone CLAUDE.md
Claude Code instructions for shaqmughal/seekstone, covering claude.md, what this repo is, commands, the harness itself (run after npm install) and architecture.
Plonk AGENTS.md
AGENTS.md instructions for ostapondo/Plonk, covering agent rules, layout, adding a module, build & verify and code style.
cyxj-groksearch CLAUDE.md
Claude Code instructions for chenyuxiaojin/cyxj-groksearch, covering claude.md, 常用命令 and 结构与坑.