get_integrityRenvoie les empreintes SHA-256 canoniques des surfaces Agent Surface, JSON et NDJSON. Utile pour vérifier qu'un agent a ingéré un snapshot précis.
| Name | Required | Description | Default |
|---|---|---|---|
| path | No | Chemin optionnel, par exemple /api/v1/claims.ndjson. |
| Name | Required | Description | Default |
|---|---|---|---|
| count | No | ||
| error | No | ||
| version | No | ||
| algorithm | No | ||
| generated | No | ||
| snapshots | No | ||
| verification | No | ||
| canonicalization | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description goes beyond the read-only and idempotent annotations by naming exactly what is returned: canonical SHA-256 fingerprints across three surface types. It does not explain domain terms like 'Agent Surface' or what happens when path is omitted, but the annotations already establish the safety profile and no contradiction is present.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two compact sentences front-load the core output and then provide a concrete use case. Every sentence earns its place; there is no fillr or redundancy.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a read-only, idempotent tool with one optional parameter and an output schema, the description is mostly sufficient. It could be more complete by defining 'Agent Surface' and explaining the default behavior when path is omitted, but the schema and annotations cover most operational needs.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The-only optional path parameter is fully described in the schema, and the description does not add further meaning about how path selects or filters surfaces. With 100% schema description coverage, the baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb and resource: it returns canonical SHA-256 fingerprints for the Agent Surface, JSON, and NDJSON surfaces, which clearly identifies the operation. However, it does not explicitly differentiate itself from siblings like verify_artifact or get_freshness, so sibling differentiation is left to inference.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It states a concrete use case: verifying that an agent ingested a precise snapshot. It does not, however, say when to prefer this tool over alternatives such as verify_artifact or get_changefeed, nor does it provide any exclusions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Most tools target clearly distinct resources (articles, claims, sources, indices, signal history, integrity), and the descriptions are detailed. The two real ambiguities are get_changefeed vs get_changes, which are near-identical except for filtering, and the Agent Surface metadata cluster (manifest, openapi, integrity, verify, ndjson feed) where an agent could misselect. These are minor enough not to undermine the set.
All 21 tools follow a consistent verb-first snake_case pattern with clear semantic verbs: get_ for retrieval, list_ for enumeration, search_ for querying, plus build_research_pack and verify_artifact as lone but clearly communicative composites. There is no mixing of conventions or vague verbs like process or run, making tool selection predictable.
21 tools is above the ideal 3-15 range, but the server covers a broad domain: content retrieval, claims and evidence graphs, risk signals, change monitoring, and integrity verification. A few tools could be consolidated (the changefeed pair and the metadata cluster), but none is pure filler, so the count feels justified though slightly heavy.
For a read-only intelligence platform, the coverage is comprehensive: content listing and full-text retrieval, claims and evidence graphs, source auditing, risk indices with history, freshness and risk-diff monitoring, and artifact verification. Minor gaps include no per-source detail endpoint and no push or subscription mechanism, but agents can work around these using list_sources and the changefeed.