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 cyanheads/openstreetmap-mcp-server --skill openstreetmap-guidegit clone --depth 1 https://github.com/cyanheads/openstreetmap-mcp-serverWrote 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/cyanheads/openstreetmap-mcp-server/openstreetmap-guide)<a href="https://agentmods.dev/skills/cyanheads/openstreetmap-mcp-server/openstreetmap-guide"><img src="https://agentmods.dev/badge/skills/cyanheads/openstreetmap-mcp-server/openstreetmap-guide/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/cyanheads/openstreetmap-mcp-server/openstreetmap-guide"><img src="https://agentmods.dev/badge/skills/cyanheads/openstreetmap-mcp-server/openstreetmap-guide.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 | $0.00050 | $0.01055 |
| Opus 5.5 | $0.00020 | $0.00422 |
| Sonnet 5.5 | $0.00010 | $0.00211 |
| Haiku 4.5 | $0.00005 | $0.00105 |
Grade A, and why
openstreetmap-guide 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 today.
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 — 50 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Scope
Use OpenStreetMap to resolve locations and investigate mapped features. Nominatim ranks place/address matches; Overpass selects features by tags and geography. A survey describes mapped objects matching the query, not a complete inventory of real-world facilities.
Identifiers
- Geocoding tools return
osm_typeandosm_idwhen available. Combine them as N + node ID, W + way ID, or R + relation ID foropenstreetmap_lookup_objects. Raw Overpass elements usetypeandidfor the same identity. - Nominatim
place_idis internal to its instance; never use it as an OSM object ID or carry it between deployments. - Boundary references come from the selected object's OSM identity. Only a relation or closed way can supply area scope; a node cannot. A W/R prefix alone does not establish that Overpass has an area for the object.
Gates
- Establish the intended scope. Inspect the resolved name, address, and object type before using its coordinates or boundary. Reverse geocoding selects a nearby indexed object; it does not establish which polygon contains the point. A place's bounding box can extend beyond its boundary polygon.
- Select tag-bearing objects before interpreting attributes. Nominatim can resolve a same-name object different from the one carrying the desired tags. An absent attribute in returned
extratagsdescribes that object, not all of OpenStreetMap. Discover by attribute with Overpass, then enrich those exact IDs if addresses are needed. - Check the filter's meaning.
amenityis one OSM key, not a generic category. Museums usetourism=museum, parksleisure=park, and supermarketsshop=supermarket. Read a category warning before concluding that a sparse result reflects sparse mapping. Exact equality differs from membership in a semicolon list such ascuisine=pizza;italian. - Check coverage before claiming absence. An empty boundary query can mean the ref mapped to no Overpass area; read its notice. An endpoint can serve only a regional extract and return an empty success outside it. Check the serving endpoint and reported freshness when results are unexpectedly empty or stale; if coverage is unknown, qualify the conclusion. An omitted timestamp supplies no evidence of freshness.
- Finish retrieval before calling the query complete. Include relations when relevant to the requested features; convenience tools otherwise search nodes and ways. Follow
nextOffsetwhile more results remain, keeping the query unchanged. Resolve deferred lookup IDs on subsequent calls. For raw results with withheld arrays, follow the returned retrieval recipe before analyzing geometry or relation membership. - Distinguish a lookup miss from a missing OSM object.
not_foundmeans no Nominatim record. Use the raw existence query in the response notice when existence matters, and apply the same coverage/freshness checks before interpreting its result.
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.
- today First seen · 50 lines · 50 tokens per session scan A 79d1d8f31854
openstreetmap-guide is a skill published in the GitHub repository cyanheads/openstreetmap-mcp-server (8 stars, last pushed yesterday), licensed Apache-2.0. It adds 50 tokens to every session and 1,055 once invoked, about $0.0002 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 2026-10-27.
Other skills, from other repositories
google-maps
Search places, resolve addresses, compare routes, inspect neighborhoods, and retrieve geographic or environmental facts through the standalone @cablate/mcp-google-map CLI. Use for concrete location questions; use the travel-planning or local-seo Skill for those specialized outcomes.
google-maps-local-seo
Analyze Google Business Profile visibility, geographic grid rankings, local search keywords, and nearby competitors through the standalone Google Maps CLI. Use for local SEO or location-strategy decisions, not ordinary place recommendations.
google-maps-travel-planning
Build evidence-backed day trips, multi-day itineraries, and efficient stop sequences with Google Maps data through the standalone CLI. Use when the desired output is a practical travel plan, not merely a route or list of places.
search
Hybrid search across Captain Memo's local memory + skills + observations. Use when the user types /captain-memo:search to retrieve top hits without the model having to decide whether to call searchall on its own.
doctor
Health probe across the worker, embedder, summarizer, queue, vector index, schema, cross-AI wiring and plugin install (cache and live versions). Use when the user types /captain-memo:doctor or when something seems wrong.
wtf-happened
Retrieve a distilled timeline of the current troubleshooting incident and generate a runbook skeleton. Returns a concise summary by default, or the full untruncated timeline on request.