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/guimatheus92/social-knowledge-base/claude-mdgit clone --depth 1 https://github.com/guimatheus92/social-knowledge-baseWrote 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/guimatheus92/social-knowledge-base/claude-md)<a href="https://agentmods.dev/instructions/guimatheus92/social-knowledge-base/claude-md"><img src="https://agentmods.dev/badge/instructions/guimatheus92/social-knowledge-base/claude-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 | $0.02490 | $0.02490 |
| Opus 5 | $0.01245 | $0.01245 |
| Sonnet 5 | $0.00498 | $0.00498 |
| Haiku 4.5 | $0.00249 | $0.00249 |
Grade A, and why
social-knowledge-base CLAUDE.md scanned grade A 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.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
There is a **Next.js 16 + TypeScript + Tailwind 4 + shadcn/ui (Base UI)** app in [`app/`](app/) that gives a UI to the download: add multiple accounts, choose media, hit Play, watch live progress (SSE), size, time, brows How it starts
The opening of the file, as written. The whole thing — 90 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md — social-knowledge-base
Instructions for Claude Code in this repo. Read before acting.
Project
Knowledge base built from Instagram videos (Reels, Stories, and Highlights) of one or more accounts. Opus watches (frames + OCR of what appears on screen) and listens (PT-BR transcription) to each video and records what it learned in markdown.
Goal
Download ALL videos from an account (including "Highlights"), analyze each one, and turn the content into queryable knowledge — for many videos, in a resumable way (pause and continue without reprocessing).
How it works (4 layers)
- Download (Python / gallery-dl) —
scripts/download_instagram.py(or the app inapp/) downloads the videos todownloads/<profile>/<tab>/and records progress. Authenticates via a browsercookies.txt(instaloader is dead: its GraphQL web endpoint returns401since ~Jan/2025). Injectsffmpeginto PATH (otherwise the video comes out without audio). - Bulk transcription (GPU) —
scripts/transcribe_gpu.pyruns faster-whispermediumon CUDA (~16–24× real time) with a domain glossary, and writes sidecars:downloads/<profile>/transcripts/<id>.{txt,json}+ an<id>.vttnext to the video. This is the step that scales (~9.6k videos in hours, not the ~150h on CPU). - Notes (Claude Code + Opus + MCP
video-analyzer) — followingprompts/build-notes.md, processes 1 video per iteration: reads the transcription (the MCP reuses the.vttas a sidecar, without re-running Whisper) and watches frames + OCR via MCP, writingnotes/<profile>/videos/<id>.md. - Synthesis + RAG —
prompts/synthesize-overview.mdgeneratesnotes/<profile>/OVERVIEW.md(theme-by-theme summary);scripts/index_transcripts.py+scripts/index_notes.py+scripts/query.pyprovide search (the library becomes searchable with the transcription alone, even before the curated note). Script details:scripts/CLAUDE.md.
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 · 90 lines · 2,490 tokens per session scan A 81db8fc3bc3f
social-knowledge-base CLAUDE.md is an instructions file published in the GitHub repository guimatheus92/social-knowledge-base (6 stars, last pushed 2mo ago), licensed MIT. It adds 2,490 tokens to every session, about $0.0124 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
llm-wiki-agent GEMINI.md
Gemini CLI instructions for SamurAIGPT/llm-wiki-agent, covering llm wiki agent — schema & workflow instructions, how to use, directory layout, page format and ingest workflow.
OpenKB AGENTS.md
AGENTS.md instructions for VectifyAI/OpenKB, covering agents.md — openkb map for coding agents, read next, dev commands, module map (openkb/) and hard invariants.
mcp-devtools CLAUDE.md
Claude Code instructions for sammcj/mcp-devtools, covering claude.md, commands, build and run, testing and code quality.
kaas CLAUDE.md
Claude Code instructions for bybit-exchange/kaas, covering claude.md, behavioral guidelines, 1. think before coding, 2. simplicity first and 3. surgical changes.
chimera-agent AGENTS.md
AGENTS.md instructions for brcampidelli/chimera-agent, covering working in this repository, where the rules live, hard rules, verifying and python — the whole suite.
goal-flight AGENTS.md
AGENTS.md instructions for simonrowland/goal-flight, covering agent notes, goal flight routing, project commands (this repo), git workflow (this repo) and git-visible trigger hygiene.