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 fmind/dot --skill api-clientgit clone --depth 1 https://github.com/fmind/dotWrote 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/fmind/dot/api-client)<a href="https://agentmods.dev/skills/fmind/dot/api-client"><img src="https://agentmods.dev/badge/skills/fmind/dot/api-client/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/fmind/dot/api-client"><img src="https://agentmods.dev/badge/skills/fmind/dot/api-client.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 | $0.00016 | $0.00770 |
| Opus 5.5 | $0.00006 | $0.00308 |
| Sonnet 5.5 | $0.00003 | $0.00154 |
| Haiku 4.5 | $0.00002 | $0.00077 |
Grade A, and why
api-client 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 8d 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 — 38 lines — stays where its author put it; the contents beside it link to each section on GitHub.
API Client
Integrate the needed API operation with a small, testable boundary. Prefer an existing CLI or maintained SDK when it satisfies the contract; xh owns endpoint inspection and provider skills own their existing clients.
Workflow
- Choose the seam: reuse the project's client and sync/async model. Otherwise use HTTPX through
uvwith one owned client lifetime, explicit timeouts and pool limits, and typed request/response boundaries. Avoid a generic SDK framework for one endpoint. - Bound the operation: cap pages, records, response bytes, concurrency, attempts, and elapsed time according to configuration. HTTPX read timeouts bound inactivity, not the total duration; enforce the operation's overall deadline separately.
- Control traversal: stop on terminal or repeated cursors, and report partial results explicitly when a limit is reached. Validate provider-returned next URLs before following them; do not send credentials to a new origin or follow redirects implicitly.
- Retry by meaning: retry only documented transient failures within the total deadline. Honor valid
Retry-After; if waiting would exceed the remaining deadline, report exhaustion instead of retrying early. Otherwise use bounded backoff with jitter. Keep authentication, permission, validation, and malformed-success failures visible. - Reconcile writes: reuse a stable provider-supported idempotency key for the same logical write. After a timeout or lost response, read authoritative operation state before retrying; without safe reconciliation or documented idempotency, return an explicit unknown outcome.
- Verify offline first: exercise the HTTPX test boundary, including limits and failures; then perform only the live reads or writes authorized for this integration. Poll asynchronous jobs with a bounded deadline and verify their terminal result.
Gotchas
- Transport retries skip status codes: HTTPX transport retries cover only failed connects (
ConnectError,ConnectTimeout); they do not implement status-code retries or safe replay of an uncertain write. - Bound streams while reading: enforce byte limits while reading and always close responses and clients, including cancellation paths. Checking size after buffering is not a memory bound.
- Route credentials deliberately: keep TLS verification enabled and make proxy/environment behavior deliberate. Validate base URLs at the configuration boundary; log only redacted endpoint identities.
- Never disguise errors as empty: an empty list must mean no records, not a swallowed 403, exhausted page budget, or failed parse. Preserve the original exception cause without including secrets.
What ships with it
1 file 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.
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.
- 8d ago Changed · -2 lines 49f0b83f9cbc
- 10d ago Changed 267253999406
- 11d ago Changed · -3 tokens per session 8a6ab360273b
- 29d ago Changed · +1 lines · -10 tokens per session 2488cfc8adf0
- 1mo ago Changed · +1 lines 9cf50e919f4a
- 1mo ago First seen · 38 lines · 29 tokens per session scan A 2a674446dd29
api-client is a skill published in the GitHub repository fmind/dot (11 stars, last pushed 2d ago), licensed MIT. It adds 16 tokens to every session and 770 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-09-10.
Other skills, from other repositories
python-stack
Build typed Python projects with uv, Ruff, ty, pytest, Litestar, and Typer. Use for packages, CLIs, web apps, tests, typing, or API verification.
python-script
Write standalone single-file Python scripts using PEP 723 inline metadata and uv. Use when creating a quick CLI script that needs dependencies without a full project.
scaffold-python
Scaffold a complete Python project with CI/CD, release pipeline, justfile, sr.yaml, pyproject.toml, .envrc, and standard files. Uses uv, ruff, and justfile (Python lacks a native task runner like pnpm scripts, so just fills that gap). Loads on top of scaffold-project (run that first for cross-language standard files).…
go-stack
Build Go projects, libraries, CLIs, TUIs, web apps, or ADK agents with the standard package layout and pinned tooling.
turbodrf
Use when building Django REST APIs with TurboDRF - model Meta configuration, role-based and field-level permissions, multi-tenant predicates, router wiring, management commands, settings, or troubleshooting.
aiohttp
Use when building Python async HTTP services or clients with aiohttp - web server routing, middleware, WebSocket, SSE, streaming, client sessions, pytest-aiohttp testing, or troubleshooting SSL and timeout issues.