Skip to main content
Glama
skyphusion-labs

@skyphusion/common-thread-mcp

Official

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
COMMON_THREAD_API_URLNoBackend base URL. Defaults to the hosted backend.https://common-thread-backend.skyphusion.org
COMMON_THREAD_ACCESS_TOKENNoDefault access token for tools. Can be overridden per tool call.
COMMON_THREAD_CF_AIG_TOKENNoCloudflare AI Gateway token (alternative to using COMMON_THREAD_AI_GATEWAY_URL + COMMON_THREAD_ANTHROPIC_API_KEY) for BYOK attribution.
COMMON_THREAD_AI_GATEWAY_URLNoCloudflare AI Gateway URL for BYOK attribution.
COMMON_THREAD_INVESTIGATION_IDNoDefault investigation ID for tools. Can be overridden per tool call.
COMMON_THREAD_ANTHROPIC_API_KEYNoAnthropic API key for BYOK attribution.

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

CapabilityDetails
tools
{
  "listChanged": true
}

Tools

Functions exposed to the LLM to take actions

NameDescription
healthA

Backend health check (GET /). Confirms API URL, version, and hosted-API notice.

create_investigationA

Create a new investigation. Returns access_token ONCE -- store it; it cannot be recovered. New investigations are encrypted at rest; the key is derived from the token.

get_investigationB

Fetch investigation metadata and practitioner metadata.

update_investigation_metadataA

PATCH practitioner metadata: triggering_events and/or time_bounds (paper §4.2.2 / §5.2.1). Active only.

seal_investigationA

Mark investigation read-only (sealed). Ingest and attribution disabled; data stays readable with the token.

delete_investigationA

Hard-delete an ACTIVE investigation (MySQL rows + R2 investigation prefix). Sealed/archived refuse. sha256 blobs retained.

investigation_summaryC

Active seed count and manifest artifact count.

list_seedsC

List seed accounts for the investigation.

add_seedB

Add a seed account (platform + handle + basis_statement). Active only. Cap: MAX_SEED_ACCOUNTS.

remove_seedA

Soft-delete an active seed (audit row retained).

list_featuresB

Query extracted features (account / pair / event) after ingest+extract.

ingest_apify_twitterA

Upload Apify Twitter export JSON (array or {items|data}). Archives + runs extractors. Returns jobId; poll get_ingest_job until completed. Active only.

get_ingest_jobA

Poll ingest job status by job_id from ingest_apify_twitter.

attributeB

Run attribution over active seed pairs. Public host requires BYOK (ai_gateway_url + anthropic_api_key or cf_aig_token). May return 200 sync or 202 async jobId.

get_attribution_jobC

Poll async attribution job from attribute (202 response).

list_runsB

List attribution runs (summaries) for the investigation.

get_runA

Single attribution run with parsed output (claims, alternatives, declined pairs, triage).

get_packetA

Evidence packet (§8.1). Default JSON for latest run; pass run_id for a specific run. format=markdown returns markdown text; format=pdf returns base64 PDF (needs PDF worker).

list_manifestB

List archive manifest entries for the investigation.

list_signaturesC

List manifest signature records.

verify_manifestC

Verify manifest signatures for the investigation.

debug_ingestC

Dev visibility: extractor vs manifest for the investigation (not a methodology deliverable).

debug_manifestC

Dev visibility: raw manifest breakdown.

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

TDQS

B3.2/5.0

Scored across 23 tools

Disambiguation5/5

Each tool targets a distinct resource or action within the investigation lifecycle, seed management, ingest, attribution, and manifest verification. The async job polling pairs (ingest_apify_twitter/get_ingest_job, attribute/get_attribution_job) are clearly separated. No two tools have overlapping responsibilities.

Naming Consistency4/5

Most tools follow a consistent verb_noun pattern (list_, create_, get_, update_, delete_, add_, remove_, verify_, debug_). Minor deviations include 'health', 'investigation_summary', and 'attribute', which lack the standard verb prefix or object, but overall the naming is predictable.

Tool Count3/5

23 tools is at the heavy end for a single server, covering a full investigation pipeline with ingest, attribution, and manifest verification. While each tool has a specific purpose, the sheer number may overwhelm agents and suggests some tools (e.g., debug_* utilities) could be split into a separate server.

Completeness4/5

The tool surface covers the core workflow: create investigation, manage seeds, ingest data, run attribution, view runs, retrieve packets, and verify manifests. Missing features like listing all investigations or updating seeds are minor and can be worked around.

Maintenance

ActivityMaintained
ResponsivenessUnresponsive