How real projects summarise themselves for language models.
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
MarcoPorcellatoReads credentialsllms.txtMarcoPorcellato/matryca-plumber/llms.txt
aliases like `[42]` without a page still fail with `ok: false`. Spec: `docs/openspec/agent-ax-robustness.md` ### 2.4 Security & Sandbox (v1.9.9+) Graph reads and writes stay inside **`LOGSEQ_GRAPH_PATH`**. v1.9.9 adds defense
stabganReads credentialsllms.txtstabgan/openrouter-mcp-multimodal/llms.txt
IMAGE_INLINE_MAX_BYTES`, `OPENROUTER_AUDIO_INLINE_MAX_BYTES`, `OPENROUTER_VIDEO_INLINE_MAX_BYTES`. ## Security - Input/output path sandboxes (`OPENROUTER_INPUT_DIR`, `OPENROUTER_OUTPUT_DIR`); bypass `OPENROUTER_ALLOW_UNSAFE_PATHS
2389-researchllms.txt2389-research/claude-plugins/docs/llms.txt
truth - [fresh-eyes-review](https://skills.2389.ai/plugins/fresh-eyes-review/): Mandatory final sanity check before commits/PRs - catches security vulnerabilities, logic errors, and bugs that slip through tests - [documentation-audit](https://skills.2389.ai/plugins/documentation-audit/): Verify
Arthurdwllms.txtArthurdw/rovo/llms-full.txt
span several lines. The status must also appear in `# Responses`. - `# Metadata`: `@tag ` (repeatable), `@security ` (repeatable; the scheme must be registered in `api.components.security_schemes`), `@id ` (alphanumerics and `_` only; defaults
Arthurdwllms.txtArthurdw/rovo/llms.txt
span several lines. The status must also appear in `# Responses`. - `# Metadata`: `@tag ` (repeatable), `@security ` (repeatable; the scheme must be registered in `api.components.security_schemes`), `@id ` (alphanumerics and `_` only; defaults
eventscheduleReads credentialsllms.txteventschedule/eventschedule/public/llms-full.txt
user": { "id": "xR3kL9", "name": "Jane Smith", "email": "user@example.com" } } } ``` Store the `api_key` securely. It expires after one year. ### Step 3: Create a Schedule ```bash curl -X POST https://eventschedule.com/api/schedules
permawebReads credentialsllms.txtpermaweb/HyperBEAM/docs/llms-full.txt
gain high confidence that the computation was performed correctly and confidentially within the secure environment, untampered by the node operator. **Obtaining Attested Responses:** This usually involves interacting with nodes specifically
nirholasReads credentialsllms.txtnirholas/crypto-vision/llms-full.txt
Jupiter, DePin Scan, etc.) - **58 specialized AI agents** for DeFi strategy, trading, portfolio management, security analysis, and education - **Real-time WebSocket feeds** for prices, Bitcoin mempool, and DEX trades - **Anomaly
runxhqllms.txtrunxhq/runx/llms.txt
docs/publishing.md) - [Releasing](docs/releasing.md) - [Operator Console](docs/operator-console.md) - [Skill Author Runtime Contract](docs/skill-author-runtime-contract.md) - [Skill Catalog](docs/skill-catalog.md) - [Security Authority Proof](docs/security-authority-proof.md) - [Runx System Architecture](docs/architecture/runx-system.md) - [Act Model](docs/act-model.md) - [How We Test](docs/how-we-test.md) ## Examples
dykyi-romanllms.txtdykyi-roman/awesome-claude-code/llms.txt
patterns - CI/CD: pipeline, tools, deployment - Docker: core, multistage, base-images, php-extensions, compose, networking, security, buildkit, production, troubleshooting, orchestration, scanning - Frameworks: Symfony, Laravel, Yii, CodeIgniter, No-Framework (pure PHP) - Task
supernovae-stllms.txtsupernovae-st/nika/llms.txt
answer. 0.124 is the current release line. Work stays focused on maintenance, security and the media capabilities workflows need; native video is planned for 0.125. There is no hosted service
danoszzllms.txtdanoszz/penthera/site/llms.txt
Penthera > Penthera is a free, open-source security scanner and fixer for your app, and it checks your AI setup. Code scans run offline. Deterministic rules decide every finding; models
rasimmellms.txtrasimme/FlowBoard/llms.txt
backoff and session-safe routing - [Telegram Mini App](docs/concepts/telegram-mini-app.md) — ordered multi-bot HMAC auth; secure sessions; tunnel access ## Decisions (ADRs) - [ADR-0001](docs/adr/0001-live-inject-bootstrap.md) — Project context delivered via live-inject
arthjeanllms.txtarthjean/paneflow/llms.txt
install paths, core workflows, safety model. - [Architecture](ARCHITECTURE.md) - runtime architecture, threading, terminal pipeline, IPC, security posture. - [Agent Instructions](AGENTS.md) - repository rules for coding agents working in this codebase. - [MCP Bridge
major7appsllms.txtmajor7apps/pensyve/llms.txt
service are unavailable. The open-source project is in maintenance mode, with security fixes and dependency updates but no new features. The Rust engine stores memories in SQLite by default
ParthibanRajasekaranllms.txtParthibanRajasekaran/text-from-image/public/llms.txt
self-hosted. English language data ships with the package. - Network: the site's Content Security Policy limits connections to its own origin. The SDK, command line and MCP server make
backenlyllms.txtbackenly/backenly/public/llms.txt
list of problems on its own (a missing index, a table without row-level security, a missing realtime trigger), verifying each fix. Anything else becomes a finding with evidence
escapeboyllms.txtescapeboy/claude-code-kit/llms.txt
keeps a `## Tasks` checklist it reads - [Subagents](https://github.com/escapeboy/claude-code-kit/tree/master/10-subagents/examples): 8 agent definitions → `~/.claude/agents/` - [Security hooks](https://github.com/escapeboy/claude-code-kit/tree/master/13-security-hardening/hooks): PreToolUse guards with tests - [Productivity hooks](https://github.com/escapeboy/claude-code-kit/tree/master/01-global-optimization/hooks): shell-habits
sfrangulovPipes a download into a shellllms.txtsfrangulov/claude-code-handbook-ru/llms-full.txt
песочнице, который можно настроить под свой стек через `/customize`. Разбор практик — [Using LLMs to «secure source code»](https://claude.com/blog/using-llms-to-secure-source-code) (копия в репозитории: [`docs/blog-post.md`](https://github.com/anthropics/defending-code-reference-harness/blob/main/docs/blog-post.md)). Репозиторий посвящён безопасности
sfrangulovllms.txtsfrangulov/claude-code-handbook-ru/llms.txt
main, whitelist доменов для curl - [Skills от Trail of Bits](https://github.com/trailofbits/skills): security-скиллы — статический анализ через CodeQL/Semgrep, аудит кода, поиск уязвимостей ## Optional - [Telegram-канал @cc_consultant](https://t.me
A markdown file at a website's root that gives language models a short guide to the site and links to the pages worth reading.
The same idea with the content included, so an agent can read the documentation in one request.
Agents and tools that fetch documentation on someone's behalf. It's a proposal, not a standard, and support varies.
Start with a one-line summary, then sections of links with a sentence each. The examples here show what real projects do.
