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/viperrcrypto/siftly/claude-mdgit clone --depth 1 https://github.com/viperrcrypto/SiftlyWrote 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/viperrcrypto/siftly/claude-md)<a href="https://agentmods.dev/instructions/viperrcrypto/siftly/claude-md"><img src="https://agentmods.dev/badge/instructions/viperrcrypto/siftly/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 | $0.01079 | $0.01079 |
| Opus 5 | $0.00540 | $0.00540 |
| Sonnet 5 | $0.00216 | $0.00216 |
| Haiku 4.5 | $0.00108 | $0.00108 |
Grade A, and why
Siftly CLAUDE.md 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 4d 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 — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Siftly
Self-hosted Twitter/X bookmark manager with AI-powered categorization, search, and visualization.
Quick Setup
./start.sh # installs deps, sets up DB, opens browser
Or manually:
npm install
npx prisma generate && npx prisma db push
npx next dev
App runs at http://localhost:3000
AI Authentication — No API Key Needed
If the user is signed into Claude Code CLI, Siftly uses their Claude subscription automatically. No API key configuration required.
How it works:
lib/claude-cli-auth.tsreads the OAuth token from the macOS keychain (Claude Code-credentials)- Uses
authToken+anthropic-beta: oauth-2025-04-20header in the Anthropic SDK - Falls back to: DB-saved API key →
ANTHROPIC_API_KEYenv var → local proxy
To verify it's working, hit: GET /api/settings/cli-status
Key Commands
npx next dev # Start dev server (port 3000)
npx tsc --noEmit # Type check
npx prisma studio # Database GUI
npx prisma db push # Apply schema changes to DB
npm run build # Production build
Project Structure
app/
api/
categorize/ # 4-stage AI pipeline (start/stop/status via SSE)
import/ # Bookmark JSON import + dedup
search/ai/ # FTS5 + Claude semantic search
settings/
cli-status/ # GET — returns Claude CLI auth status
test/ # POST — validates API key or CLI auth
analyze/images/ # Vision analysis progress + trigger
bookmarks/ # CRUD + filtering
categories/ # Category management
mindmap/ # Graph data
stats/ # Dashboard counts
import/ # 3-step import UI
mindmap/ # Interactive force graph
settings/ # API keys, model selection
ai-search/ # Natural language search UI
bookmarks/ # Browse + filter UI
categorize/ # Pipeline monitor
lib/
claude-cli-auth.ts # Claude CLI OAuth session (macOS keychain)
categorizer.ts # AI categorization + default categories
vision-analyzer.ts # Image vision + semantic tagging
fts.ts # SQLite FTS5 full-text search
rawjson-extractor.ts # Entity extraction from tweet JSON
parser.ts # Multi-format bookmark JSON parser
exporter.ts # CSV / JSON / ZIP export
prisma/schema.prisma # SQLite schema (Bookmark, Category, MediaItem, Setting, ImportJob)
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.
- 4d ago First seen · 117 lines · 1,079 tokens per session scan A a41301491e87
Siftly CLAUDE.md is an instructions file published in the GitHub repository viperrcrypto/Siftly (3,025 stars, last pushed 4mo ago), licensed MIT. It adds 1,079 tokens to every session, about $0.0054 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 2026-08-30.
Other instructions, from other repositories
fieldtheory-cli CLAUDE.md
Instructions for afar1/fieldtheory-cli, covering claude.md, commands, architecture, key files and data flow.
karakeep AGENTS.md
AGENTS.md instructions for karakeep-app/karakeep, covering karakeep project overview, project overview, project structure, applications (apps/) and packages (packages/).
karakeep CLAUDE.md
Claude Code instructions for karakeep-app/karakeep, a project described as: A self-hostable bookmark-everything app (links, notes and images) with AI-based automatic tagging and full text search.
karakeep GEMINI.md
Gemini CLI instructions for karakeep-app/karakeep, a project described as: A self-hostable bookmark-everything app (links, notes and images) with AI-based automatic tagging and full text search.
adhx CLAUDE.md
Instructions for itsmemeworks/adhx, covering claude.md, local configuration, example claude.local.md, github cli and sentry cli.
adhx AGENTS.md
Instructions for itsmemeworks/adhx, covering agents.md, what this project is, how to get context (read in this order), context protocol (keep the loop alive) and non-negotiable invariants (details in claude.md).