VeriSwarm MCP Server

Trust infrastructure for AI agents via Model Context Protocol.

Quick Start

pip install veriswarm-mcp
veriswarm-setup

The setup wizard configures everything for your platform (Claude Code, Gemini CLI, or Codex):

MCP Server — 90+ tools spanning trust scoring, Guard security, Passport identity (including JIT access grants), Vault audit, compliance (OWASP/EU AI Act/NIST/ISO 42001), Cedar policies, ABAC agent attributes, SRE (circuit breakers + SLOs), context governance, cross-model verification, content provenance (EU AI Act Art. 50), and A2A transport keys
Guard Hooks — Automatic PII protection for prompts and tool calls
Guard Proxy — Transparent MCP interception for any tool server (optional)

What the hooks do

Hook	Event	Action
Prompt Guard	`UserPromptSubmit`	Blocks prompts containing PII (emails, SSNs, etc.)
Tool Input Guard	`PreToolUse`	Tokenizes PII in tool arguments before execution
Tool Output Guard	`PostToolUse`	Tokenizes PII in tool responses before the LLM sees them

Manual setup

If you prefer manual configuration:

{
  "mcpServers": {
    "veriswarm": {
      "command": "python",
      "args": ["-m", "veriswarm_mcp"],
      "env": {
        "VERISWARM_API_URL": "https://api.veriswarm.ai",
        "VERISWARM_API_KEY": "vs_your_platform_key"
      }
    }
  }
}

Environment Variables

Variable	Required	Description
`VERISWARM_API_URL`	No	API base URL (default: `https://api.veriswarm.ai`)
`VERISWARM_API_KEY`	Yes (or `AGENT_KEY`)	Platform API key for workspace operations
`VERISWARM_AGENT_KEY`	Yes (or `API_KEY`)	Agent-scoped key for self-reporting tools

Tools (67 total)

Auth note: Tools marked with (session) require x-account-access-token and are not available with API key only. All other tools work with x-api-key.

Trust Scoring (6)

Tool	Description
`check_trust`	Get agent's current trust scores, policy tier, and risk band
`check_decision`	Check if an agent is allowed to perform an action (allow/review/deny)
`get_my_score`	Get your own trust scores with improvement guidance (agent key)
`get_score_history`	Get score history over time for an agent
`get_score_breakdown`	Get detailed score breakdown with contributing factors
`explain_score`	Get human-readable explanation of an agent's scores

Event Reporting (4)

Tool	Description
`report_action`	Report a generic behavioral event for scoring
`report_tool_call`	Shorthand for tool.call.success / tool.call.failure events
`report_interaction`	Report an agent-to-agent interaction
`report_incident`	Report a security incident for Guard review

Guard Security (12)

Tool	Description
`scan_tool`	Request a security scan for a tool or MCP server
`scan_injection`	Scan text for prompt injection patterns
`check_tool_allowed`	Check if a tool is permitted under active Guard policies
`get_findings`	List Guard security findings, optionally filtered by agent
`list_guard_policies`	List all active Guard policies for the workspace
`kill_agent`	Activate kill switch for an agent (blocks all trust decisions)
`unkill_agent`	Deactivate kill switch, restoring normal processing
`tokenize_pii`	Remove PII from text, replacing with safe tokens
`rehydrate_pii`	Restore original PII values from VeriSwarm tokens
`get_pii_session`	Get details of a PII tokenization session
`revoke_pii_session`	Revoke a PII session, deleting all stored tokens

Passport Identity (4)

Tool	Description
`get_credentials`	Issue a signed JWT Passport credential (agent key required)
`verify_credential`	Verify a Passport JWT and return decoded trust claims
`verify_identity`	Mark an agent as identity-verified (admin action)
`check_delegation`	Check active delegation grants for an agent

Vault Audit (3)

Tool	Description
`query_ledger`	Query the immutable Vault audit ledger
`verify_chain`	Verify hash-chain integrity of the Vault ledger
`export_vault`	Create a Vault export job (JSON or CSV)

Agent Management (5)

Tool	Description
`register_agent`	Register a new agent and return its agent_id
`get_agent`	Get full agent profile including trust scores
`get_agent_timeline`	Get an agent's event timeline
`get_agent_flags`	Get active moderation flags for an agent
`list_agents`	List workspace agents with search and tier filtering

Platform & Workspace (9)

Tool	Description
`platform_status`	Check platform health, uptime, and feature flags
`get_scoring_profile`	Get workspace scoring profile and weights
`set_scoring_profile`	Set workspace scoring profile
`list_notifications`	List recent workspace notifications
`get_ip_allowlist`	Get IP allowlist configuration
`set_ip_allowlist`	Set IP allowlist CIDRs
`get_custom_domain`	Get custom domain configuration
`set_custom_domain`	Set a custom domain
`list_team_members`	List workspace team members
`invite_team_member`	Invite a new team member

Conversation Analytics (2)

Tool	Description
`get_agent_analytics`	Get aggregated quality analytics for an agent (resolution, accuracy, tone, efficiency, security)
`get_cost_recommendations`	Get cost optimization recommendations (model downgrades, KB gaps, prompt tuning)

Security Testing (3)

Tool	Description
`list_red_team_attacks`	List available adversarial attack patterns (36 attacks across 5 categories)
`check_tool_permission`	Check trust-gated permission for a tool based on agent's trust score
`score_conversation_security`	Compute per-conversation security grade (A-F) from Guard + Vault data

GDPR Compliance (3)

Tool	Description
`forget_contact`	Delete all data associated with a contact (with cryptographic deletion proof)
`list_deletion_records`	List GDPR deletion records with verification hashes
`verify_deletion`	Verify a deletion record exists and return its proof

A2A Trust (2)

Tool	Description
`get_agent_reputation`	Get cross-platform reputation score for an agent
`get_a2a_history`	Get agent-to-agent interaction history (inbound or outbound)

Benchmarking & Operations (2)

Tool	Description
`get_benchmark_history`	Get benchmark results for an agent across versions
`get_provider_health`	Get LLM provider health status (success rate, latency, cooldown)

Development

cd packages/mcp-server
pip install -e .
python -m pytest tests/ -v

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

veriswarm_mcp-0.4.0.tar.gz (36.5 kB view details)

Uploaded Jun 2, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

veriswarm_mcp-0.4.0-py3-none-any.whl (42.2 kB view details)

Uploaded Jun 2, 2026 Python 3

File details

Details for the file veriswarm_mcp-0.4.0.tar.gz.

File metadata

Download URL: veriswarm_mcp-0.4.0.tar.gz
Upload date: Jun 2, 2026
Size: 36.5 kB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for veriswarm_mcp-0.4.0.tar.gz
Algorithm	Hash digest
SHA256	`ba042e29b9c7cd1e97948f24e895f2efabed4f92dc86ee9818f18e06e13cf5fb`
MD5	`b42d167daac104546d1a82446baa8587`
BLAKE2b-256	`595a3857da1c97447a1a4d5ed230a55b474e63a1ad434fe8ca84503848ccf829`

See more details on using hashes here.

Provenance

The following attestation bundles were made for veriswarm_mcp-0.4.0.tar.gz:

Publisher: publish-mcp.yml on veriswarm/veriswarm-sdk

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: veriswarm_mcp-0.4.0.tar.gz
- Subject digest: ba042e29b9c7cd1e97948f24e895f2efabed4f92dc86ee9818f18e06e13cf5fb
- Sigstore transparency entry: 1704932565
- Sigstore integration time: Jun 2, 2026
Source repository:
- Permalink: veriswarm/veriswarm-sdk@3cbbd44d5ae7651665e12c5b98d24242cafe1da2
- Branch / Tag: refs/tags/mcp-v0.4.0
- Owner: https://github.com/veriswarm
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-mcp.yml@3cbbd44d5ae7651665e12c5b98d24242cafe1da2
- Trigger Event: push

File details

Details for the file veriswarm_mcp-0.4.0-py3-none-any.whl.

File metadata

Download URL: veriswarm_mcp-0.4.0-py3-none-any.whl
Upload date: Jun 2, 2026
Size: 42.2 kB
Tags: Python 3
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for veriswarm_mcp-0.4.0-py3-none-any.whl
Algorithm	Hash digest
SHA256	`fe15b2df678b2ec6c763c5d83dabc092476287da2ad67c4555562bbf7a518c97`
MD5	`e1e0213f96219261e3fc2d9075af7dd6`
BLAKE2b-256	`592a278d9c28e4d2944653a4c9eb62dfdf5791d50a832e702b42740dc5a0f3ce`

See more details on using hashes here.

Provenance

The following attestation bundles were made for veriswarm_mcp-0.4.0-py3-none-any.whl:

Publisher: publish-mcp.yml on veriswarm/veriswarm-sdk

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: veriswarm_mcp-0.4.0-py3-none-any.whl
- Subject digest: fe15b2df678b2ec6c763c5d83dabc092476287da2ad67c4555562bbf7a518c97
- Sigstore transparency entry: 1704932614
- Sigstore integration time: Jun 2, 2026
Source repository:
- Permalink: veriswarm/veriswarm-sdk@3cbbd44d5ae7651665e12c5b98d24242cafe1da2
- Branch / Tag: refs/tags/mcp-v0.4.0
- Owner: https://github.com/veriswarm
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish-mcp.yml@3cbbd44d5ae7651665e12c5b98d24242cafe1da2
- Trigger Event: push

Release history Release notifications | RSS feed

This release

0.4.0 This release

Jun 2, 2026 2 files

0.3.1

Jun 1, 2026 2 files

0.3.0

May 20, 2026 2 files

0.2.0

May 12, 2026 2 files

veriswarm-mcp 0.4.0

VeriSwarm MCP Server

Quick Start

What the hooks do

Manual setup

Environment Variables

Tools (67 total)

Trust Scoring (6)

Event Reporting (4)

Guard Security (12)

Passport Identity (4)

Vault Audit (3)

Agent Management (5)

Platform & Workspace (9)

Conversation Analytics (2)

Security Testing (3)

GDPR Compliance (3)

A2A Trust (2)

Benchmarking & Operations (2)

Development

Project links

Key dates

1 maintainer

Credits

GitHub Statistics

License

Requires

Tags

Classifiers

Download files

Source Distribution

Built Distribution

File details

File metadata

File hashes

Provenance

File details

File metadata

File hashes

Provenance

Release history Release notifications | RSS feed