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 skills add ngerakines/atproto-skills --skill atproto-lexicongit clone --depth 1 https://github.com/ngerakines/atproto-skillsWrote 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/ngerakines/atproto-skills/atproto-lexicon)<a href="https://agentmods.dev/skills/ngerakines/atproto-skills/atproto-lexicon"><img src="https://agentmods.dev/badge/skills/ngerakines/atproto-skills/atproto-lexicon/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/ngerakines/atproto-skills/atproto-lexicon"><img src="https://agentmods.dev/badge/skills/ngerakines/atproto-skills/atproto-lexicon.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00720 | $0.03527 |
| Opus 5 | $0.00360 | $0.01764 |
| Sonnet 5 | $0.00144 | $0.00705 |
| Haiku 4.5 | $0.00072 | $0.00353 |
Grade A, and why
atproto-lexicon 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 12d 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 — 144 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AT Protocol Lexicon + XRPC + Records
An AT Protocol lexicon is a JSON document that declares the shape of one record type or one XRPC method. Every record's $type, every XRPC path, and every ref between them ties back to an NSID. The lexicon is the protocol's type system; validation against it is the contract between clients, servers, and the repository layer.
This skill routes to per-language guides for Rust, TypeScript, and Go, sitting on top of a language-neutral spec in shared/.
Defaults
- Lexicon doc —
{lexicon: 1, id: <nsid>, revision?, defs: {...}}. Primary types (record,query,procedure,subscription) live undermain. Secondary types can use any name. - NSID — reversed-DNS authority + name segment. ASCII only, 317 bytes max, name segment allows no hyphens. See
shared/nsid.md. $type— required on every record. Bare NSID implies#main. Used for union dispatch. Missing$type= invalid record.- AT-URIs in records — DIDs strongly preferred over handles for the authority. Query strings forbidden. See
shared/at-uri.md. - strongRef —
{uri, cid}wherecidis a plain string, not acid-link. Common bug source. - Blob refs (modern) —
{$type:"blob", ref:{$link:<cid>}, mimeType, size}. Legacy{cid, mimeType}only accepted with an opt-in lenient flag. - XRPC wire —
/xrpc/<nsid>path. GET for queries, POST for procedures, WebSocket for subscriptions. Errors as{error, message}. Subscription frames are two concatenated DAG-CBOR objects per WS binary message. - Validation — strict by default on write, lenient on read.
ALLOW_LEGACY_BLOBandALLOW_LENIENT_DATETIMEare the typical lenient opt-ins.
Full normative rules: shared/lexicon-spec.md, shared/nsid.md, shared/at-uri.md, shared/record-model.md, shared/xrpc-wire.md, shared/backward-compat.md. Fixtures: shared/test-vectors.md. Cross-language differences: shared/divergence-matrix.md.
Language detection
Before generating or reviewing any lexicon / XRPC / record code, determine the target language from project files or the file being edited:
What ships with it
23 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- go/authoring.md 5.7 KB
- go/README.md 8.6 KB
- go/records.md 6.6 KB
- go/validation.md 5.7 KB
- go/xrpc-client.md 7.1 KB
- rust/authoring.md 5.3 KB
- rust/README.md 7.8 KB
- rust/records.md 6.5 KB
- rust/validation.md 7.7 KB
- rust/xrpc-client.md 7.1 KB
- shared/at-uri.md 4.6 KB
- shared/backward-compat.md 6.6 KB
- shared/divergence-matrix.md 9.7 KB
- shared/lexicon-spec.md 9.5 KB
- shared/nsid.md 3.5 KB
- shared/record-model.md 6.0 KB
- shared/test-vectors.md 7.7 KB
- shared/xrpc-wire.md 5.1 KB
- typescript/authoring.md 4.7 KB
- typescript/README.md 6.6 KB
- typescript/records.md 5.1 KB
- typescript/validation.md 5.9 KB
- typescript/xrpc-client.md 6.9 KB
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.
- 12d ago First seen · 144 lines · 720 tokens per session scan A f5f5ba4e7b81
atproto-lexicon is a skill published in the GitHub repository ngerakines/atproto-skills (50 stars, last pushed 4mo ago), licensed MIT. It adds 720 tokens to every session and 3,527 once invoked, about $0.0036 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 skills, from other repositories
afrexai-api-architect
Design, build, test, document, and secure production-grade APIs. Covers the full lifecycle from schema design through deployment, monitoring, and versioning. Use when designing new APIs, reviewing existing ones, generating OpenAPI specs, building test suites, or debugging production issues.
litellm
Operate, configure, secure, and troubleshoot the LiteLLM AI gateway (proxy) and Python SDK: run the proxy (litellm --config), route to 100+ providers through one OpenAI-compatible API, configure model lists and routing/reliability, virtual keys, teams, budgets, rate limits, caching, guardrails, observability, and…
Send and diagnose transactional email through Twilio SendGrid from a terminal or agent: send messages, check deliverability (bounces and spam reports), and verify Signed Event Webhook signatures (ECDSA P-256) — with a bundled email-cli script that is read-only by default and gates every send behind a --dry-run/--yes…
supabase
Use this skill when developing applications with Supabase, running the Supabase CLI, designing migrations and RLS policies, testing database behavior, generating client types, deploying the official self-hosted Docker stack, or administering its Postgres, Auth, Storage, Realtime, Functions, API gateway, backups…
vercel-eve
Build, develop, deploy, self-host, secure, and troubleshoot durable backend AI agents with Vercel Eve. Use when creating an Eve agent, adding tools, skills, subagents, channels, schedules, sandboxing, durable sessions, observability, or deploying Eve on Vercel or a Node host. Do not use for the separate Vercel AI SDK…
agentmail
API-first email platform designed for AI agents. Create and manage dedicated email inboxes, send and receive emails programmatically, and handle email-based workflows with webhooks and real-time events. Use when you need to set up agent email identity, send emails from agents, handle incoming email workflows, or…