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 aapiskotin/garmin-mcp --skill garmin-local-appgit clone --depth 1 https://github.com/aapiskotin/garmin-mcpWrote 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/aapiskotin/garmin-mcp/garmin-local-app)<a href="https://agentmods.dev/skills/aapiskotin/garmin-mcp/garmin-local-app"><img src="https://agentmods.dev/badge/skills/aapiskotin/garmin-mcp/garmin-local-app/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/aapiskotin/garmin-mcp/garmin-local-app"><img src="https://agentmods.dev/badge/skills/aapiskotin/garmin-mcp/garmin-local-app.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 | $0.00052 | $0.01636 |
| Opus 5.5 | $0.00021 | $0.00654 |
| Sonnet 5.5 | $0.00010 | $0.00327 |
| Haiku 4.5 | $0.00005 | $0.00164 |
Grade A, and why
garmin-local-app 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 7d 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 — 129 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Follow specs/garmin-app.md. Training analysis/advice belong to the agent.
Garmin retrieval and feature computation happen only on request. Existing Garmin
write tools still require an approved preview and confirm=true.
Paths and access
- Host data:
$GARMIN_DATA_DIR, default<checkout>/.garmin-data/. - Container data:
/data/garmin-app/; database:garmin.sqlite3. - Immutable objects:
objects/<sha256>.json, canonical raw response JSON. - Captures:
manifests/<capture_id>.json, with account, source, object hash, request boundaries/parameters, timezone, and UTC fetched timestamp. - Tokens: separate
garmin_tokensCompose volume at/data/garmin-tokens. - Features:
src/garmin_mcp/features/*.py; migrations:src/garmin_mcp/migrations/*.sql. Compose mounts both directories into the app. - Used definition code:
definitions/<code_hash>.pyin the data directory. - Backups:
backups/<backup_id>/, withbackup.jsonchecksums and code/config IDs. - Local endpoints:
http://127.0.0.1:8765/mcp,/docs,/openapi.json,/. The UI only reads cached records. Run one server process per data directory.
Use the Docker runtime for DB operations if host Python lacks patched SQLite.
The image builds verified SQLite 3.51.3; the app rejects unpatched WAL versions.
Run Python with docker compose exec garmin-mcp python; import
garmin_mcp.server.local_service for access to service.store.
Never expose credentials in features, responses, or raw-object metadata.
Schema and freshness
Read src/garmin_mcp/migrations/001_initial.sql for metadata tables and keys.
activities uses (account_id, activity_id); days uses
(account_id, local_date). Inspect PRAGMA table_info for source scalar columns
and feature_<name> columns added after initialization.
Activity local_date uses the calendar timezone when UTC time is available;
source_local_date and start_time_local preserve Garmin's original context.
sources records per-source status, capture, normalized nested metadata, original
source date, fetched time, and attempted time. captures retains superseded
responses. coverage records completed activity ranges; sync_state holds
cursors and latest errors. feature_results records versions, source references,
computed times, and statuses. Only status=ok feature values are usable; cached
SQL values can remain after invalidation. Lookup masks invalidated feature values.
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.
- 7d ago First seen · 129 lines · 52 tokens per session scan A 83f5d7de922b
garmin-local-app is a skill published in the GitHub repository aapiskotin/garmin-mcp (0 stars, last pushed 5d ago), licensed MIT. It adds 52 tokens to every session and 1,636 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 2026-10-08.
Other skills, from other repositories
qdrant-clients-sdk
Qdrant provides client SDKs for various programming languages, allowing easy integration with Qdrant deployments.
azure-cosmos-db-py
Build Azure Cosmos DB NoSQL services with Python/FastAPI following production-grade patterns. Use when implementing database client setup with dual auth (DefaultAzureCredential + emulator), service...
azure-cosmos-py
Client library for Azure Cosmos DB NoSQL API — globally distributed, multi-model database.
azure-data-tables-py
NoSQL key-value store for structured data (Azure Storage Tables or Cosmos DB Table API).
alembic
Manage database migrations for SQLAlchemy with Alembic. Use when a user asks to version database schemas, autogenerate migration scripts from models, run or roll back migrations, set up async SQLAlchemy migrations, or handle schema changes safely in production.
neo4j-driver-python-skill
Neo4j Python Driver v6 — driver lifecycle, executequery, managed and explicit transactions, async (AsyncGraphDatabase), result handling, data type mapping, error handling, UNWIND batching, connection pool tuning, and causal consistency. Use when writing Python code that connects to Neo4j via GraphDatabase.driver…