SiteGraph
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| DSA_TDB_TOKEN | No | DSA Transparency Database Research API token (EU Login + request to the DSA helpdesk) | |
| PUBPEER_DEVKEY | No | PubPeer API key (request from PubPeer) | |
| AGENT_OSINT_HOME | No | Data directory (default ~/.agent-osint): caches, indexes, evidence | |
| OPENSANCTIONS_URL | No | URL of your self-hosted yente | |
| META_GRAPH_VERSION | No | Graph API version (default v26.0) | v26.0 |
| AGENT_OSINT_CONTACT | No | Your email, sent in the User-Agent and to Crossref/OpenAlex 'polite pools' (recommended) | |
| CERTSPOTTER_API_KEY | No | SSLMate Cert Spotter key for higher CT query limits | |
| AGENT_OSINT_HOST_RPS | No | Override rate limits, e.g. api.crossref.org=10,pypi.org=5 | |
| AGENT_OSINT_CACHE_TTL | No | Cache lifetime (s) | |
| AGENT_OSINT_MAX_BYTES | No | Response cap (bytes) | |
| META_AD_LIBRARY_TOKEN | No | Meta Ad Library access token | |
| OPENSANCTIONS_API_KEY | No | OpenSanctions hosted API key | |
| AGENT_OSINT_FILE_ROOTS | No | os.pathsep-separated folders that local-file tools may read (recommended when exposing to agents) | |
| AGENT_OSINT_RECORD_DIR | No | Record live responses as fixtures | |
| AGENT_OSINT_REPLAY_DIR | No | Replay recorded fixtures offline | |
| COMPANIES_HOUSE_API_KEY | No | UK Companies House REST API key | |
| REGULATIONS_GOV_API_KEY | No | Free key from api.data.gov | |
| AGENT_OSINT_HTTP_TIMEOUT | No | Timeout (s) | |
| GOOGLE_FACTCHECK_API_KEY | No | Google Cloud key with the Fact Check Tools API enabled | |
| AGENT_OSINT_TORTURED_PHRASES | No | Extra tortured-phrase list (tortured => expected per line) | |
| AGENT_OSINT_ALLOW_PRIVATE_NETWORK | No | 1 to allow fetching private/loopback addresses (off by default) |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| claim_timelineC | Merged, dated timeline for a claim across URLs, archives and fact-checks. |
| url_timelineC | First/latest archive capture, declared publish dates (live + earliest archived copy), removal status. |
| fact_checksC | Search published fact-checks (Google Fact Check Tools API; needs GOOGLE_FACTCHECK_API_KEY). |
| claimreview_from_urlB | Extract schema.org ClaimReview data (claim, rating, dates, appearances) from a fact-check article. |
| image_provenanceB | Read C2PA Content Credentials: signer, time, actions, AI-generation declarations, validation state. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
The tools have mostly distinct scopes: claim_timeline operates on a claim, url_timeline on a URL, fact_checks searches an API, claimreview_from_url parses a specific article, and image_provenance handles C2PA data. Some overlap exists because claim_timeline aggregates across URLs and fact-checks, but the descriptions make the boundaries reasonably clear.
All tool names use lowercase snake_case and are descriptively noun-based, which is consistent enough to be predictable. Minor deviations are claimreview_from_url (prepositional suffix) and fact_checks (plural), but the overall naming style is coherent.
Five tools is well-scoped for an investigative claim/archive-provenance server, and each tool maps to a distinct research need. There is no obvious filler or missing major workflow at this size.
The surface covers claim timelines, URL archive timelines, fact-check search, ClaimReview extraction, and image provenance, which is strong coverage for the domain. Minor gaps remain, such as direct source/URL discovery or batch investigation helpers, but core workflows are supported.