Trust Gate MCP provides tools to mint and verify tamper-evident receipts, gate consequential agent actions, audit tool inventories, check egress sensitivity, and run exit drills.
Mint post-quantum receipts for CRM record changes (mint_receipt_for_record_change) using hashed old/new values.
Rank caller-provided MCP tool inventories by worst-regret risk (audit_my_agent_inventory, read-only, no receipt).
Mint receipts for arbitrary consequential agent actions (mint_action_receipt) with optional caller-claimed attestation fields.
Verify receipts offline from the certificate alone (verify_receipt), including PQ-required mode and optional signer pinning via expected_kid.
Run a two-phase decision gate (gate_decision) that previews risk and commits a verdict, returning ALLOW/DENY/ESCALATE and a signed receipt with GRANTED/DENIED/WITHHELD_PENDING_HUMAN.
Classify egress data samples for sensitivity (check_egress) as NO_MARKERS_FOUND/INTERNAL/CONFIDENTIAL/RESTRICTED and flag blocked egress.
Run a vendor exit readiness drill (run_exit_drill) that checks the signing key, local model endpoint configuration, and local data export, returning step results and a receipt.
Allows verification of local model access via Ollama as part of vendor exit readiness drills.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Trust Gate MCP mint receipt for approving high-value purchase"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Tamper-evident receipts for consequential agent actions, as an MCP server, with a decision gate that returns ALLOW, DENY or ESCALATE.
Seven tools, one shared signing primitive: the open-source OpenAgentOntology mint_receipt.
What actually gets signed depends on what you install -- OAO detects its backend at import time, so this is worth stating plainly rather than advertising the best case:
Install | Legs | Notes |
| Ed25519 + ML-DSA-65 | Default. The ML-DSA leg comes from |
| Ed25519 + ML-DSA-65 (native backend) | Installs |
PQ-required verify (the default) demands at least one verified post-quantum leg. That shows a post-quantum signature is present and valid over the same bytes. It does not say whose key made it (see "What a valid receipt proves"), and it does not make the default backend production-grade.
Tool | What it does |
| Mints a receipt for a CRM record change. Works with any CRM (open-core Relaticle, hosted CRMs via their own MCP, custom). Old/new values are SHA-256 hashes (unsalted, so a low-entropy value can be guessed from its hash). |
| Ranks a CALLER-PROVIDED list of MCP tools by worst-regret if they act (a name-based heuristic that weights a name's first word most: |
| Receipt for any consequential agent action. Attestation values ( |
| Verify a receipt from the certificate alone -- offline, no DB. Refuses unsigned receipts, checks the |
| Two-phase decision gate with a real verdict. PREVIEW returns ALLOW, DENY or ESCALATE with the risk tier and reasons, without acting. COMMIT re-evaluates and mints a signed receipt for the verdict. Only ALLOW returns a |
| Flags sensitive markers in a data sample (a finite list of keywords, SSN and card-number shapes, and common credential formats, including |
| Vendor exit readiness drill. Checks the local signing key (names the post-quantum backend and whether the hash-based leg is available) and that a local model endpoint is configured (Ollama; it is not contacted, and a remote host is not counted). Returns step-by-step results + receipt. Informational; it signs one receipt, which creates the signing key on a host that has none yet. |
Scope. It applies only to actions your code routes through gate_decision. It does not observe, intercept or block anything an agent does, and it does not check afterward whether an action happened. Your code (a hook, a wrapper) must ask first and must obey the answer.
ALLOW is a closed list. An action gets ALLOW only when its name is a known read verb (read, get, list, search, query, find, show, view, inspect, describe) followed only by words from a small built-in vocabulary of nouns (for example read_file, list_files, search_files, list_dir, get_user, describe_table), the name is a plain identifier, and the resource is built only from words of that vocabulary, a few ordinary document words (readme, guide, changelog, ...) and source or prose file types (md, txt, py, ...), with digits allowed between words (docs/2024/notes.md). A name that contains any other word escalates to a human, because the gate has not seen it. This is deliberately narrow: most real file names escalate. Widening it means adding words to the vocabulary in the source, which is a visible change.
Structural rules escalate as well. A hidden path (.env, .ssh, .git, ..), an absolute path, a URL, a host name as the first part of a path (example.com/x, localhost), a network address in any of its usual spellings (four dotted numbers of any width, hexadecimal, octal such as 0177.1, one long integer, 0; short dotted numbers such as 1.2.3 stay versions unless the last number is too large for a version or the first is 0 or 127), a resource that starts with -, a phrase of three or more words, anything containing characters other than letters, digits, spaces and _ . - / (after Unicode compatibility folding), and a file whose real type is configuration, data, key, state or archive (a rotation counter such as .1 is ignored when judging the type; a listed data or archive type earlier in a dotted name, such as users.csv.txt, counts too). A resource that carries a command escalates: a listed mutating word (alter, merge, flush, load, ...), a critical verb glued to a word (deleteall), or any multi-word resource after a query verb.
DENY is for a critical lead verb (payments, deletion, deployment, code execution, privilege changes). A risky word elsewhere in the name, or in the resource, escalates so that a human decides; so do the high-risk lead verbs (format, remove, reset, release, escalate) and a file name that contains a risky word (test_delete_user.py).
Risky and sensitive words are listed, and the lists are finite. Listed secret and regulated-data words (password, token, key, pass, config, ssn, diagnosis, ... in their glued, plural and digit-swapped forms) give the HIGH tier a specific reason. The lists are not what keeps a name from being allowed (a sensitive name on no list still escalates, because ALLOW is closed). history is not a read noun. Matching runs across word boundaries, so some harmless names show the sensitive reason (api/index.md contains "pii"). A lookup such as list_users is a read like any other: the gate does not judge how much personal data it returns. Common read tools whose names contain a word that is also a risky verb (get_email, list_messages) escalate.
It is a name-based heuristic, not a proof. ALLOW does not mean the action is safe: an agent that calls a dangerous tool read_file is not caught, because the gate sees names, not behaviour. context is recorded in the receipt as a hash and is not used to decide. There is no operator override in this package.
Read the verdict from the receipt, and check who issued it. receipt.decision (and receipt.evidence.ontology.verdict) are signed. The top-level verdict and permit fields of the response are not; verify_receipt lists the unsigned keys of any receipt it checks. A receipt is a gate verdict from your server only if verify_receipt with expected_kid set to your server's kid returns signer_pinned: true AND verified.issuer_tool is gate_decision AND verified.operation is decision_gate_commit; every other tool signs its own issuer_tool, and without the pin those fields are only claims made by whoever signed.
No permit without a receipt this server would accept. COMMIT issues GRANTED only if the receipt was minted and signed, and verifies under this server's own policy (in PQ-required mode that includes a valid post-quantum signature).
Action and resource are recorded in clear text in the signed receipt. Do not put secrets or personal data in them.
A receipt is evidence of a verdict, not a log entry. atom_id is derived from the content (40 bits of the evidence hash), so identical decisions share one and different decisions can too; key on evidence_hash and signed_at. The server keeps no log of the receipts it signs: store the ones you need. There is no replay or freshness check: compare evidence.ontology.action and resource with what you are about to do.
0.2.1 returned GRANTED for every action. Upgrade to 0.3.0. See the changelog.
Related MCP server: DCL Evaluator
verify_receipt returning ok: true means the evidence is unchanged since it was signed, the receipt carries a signature, any kid it carries matches the key it carries, and (in the default PQ-required mode) at least one post-quantum signature verified over the same bytes.
It does not mean the receipt came from a server you trust: anyone can generate keys and sign a receipt. To check who signed, pass expected_kid (the kid of the Ed25519 key you trust, learned out of band; 32 hexadecimal characters, case and surrounding spaces ignored, a blank or malformed pin is refused). A pin counts only when that key's own Ed25519 signature verifies, and signer_pinned: true says it did (a signature verifying shows that the key signed those bytes, not that anyone holds its private key: Ed25519 accepts a few special public keys that verify for any message, which cannot affect a pin on a key you trust); ok: true with signer_pinned: false is an integrity result only, and kid is reported only when the Ed25519 signature verified. A receipt that names an Ed25519 signer (key or kid) without a valid signature from it is refused: those two values are public and can be copied into a receipt signed with other keys.
On stdio the signing key is stored unencrypted under the user's home directory, so an agent running as that user can read it: a pin does not protect against the agent you are gating. The kid fingerprints the Ed25519 key only. The ML-DSA and hash-based keys inside a receipt are not covered by it, so a pinned check is only as strong against a quantum attacker as Ed25519. Results can also depend on which post-quantum backends the verifying host has installed (a leg it cannot check is reported unverifiable).
H1 key persistence + bootstrap with FAIL-CLOSED kid-drift check (container deploy; the bootstrap checks the Ed25519 key only)
H2 per-IP token-bucket rate limit (FIFO eviction). It parses the request body exactly as the MCP transport does, from the raw bytes (a byte-order mark, UTF-16 and UTF-32 are read too), and charges the tool it names; the signing tools share the signing budget. A body it cannot parse, or one over 64 KiB, is charged to the signing budget; a request is refused with 413 as soon as its body exceeds 1 MiB (at most 1 MiB is buffered). Signing (about 70 ms per call with the default backend) and verification (about 20 ms) run on the event loop and the budgets are per client, so there is no overall cap on that work. Clients are identified by the connection address as uvicorn resolves it: set TRUST_GATE_FORWARDED_ALLOW_IPS to your proxy's address, or any client can choose its own identity with an X-Forwarded-For header.
H3 PQ-required verify: refuses a receipt with no verified post-quantum signature (and any unsigned receipt)
H4 128-bit kid on every minted receipt, recomputed from the canonical key bytes and checked on verify
Optional bearer auth and a narrowed CORS allowlist via TRUST_GATE_BEARER_TOKEN + TRUST_GATE_ALLOWED_ORIGINS
The test suite pins every listed verb, noun and sensitive word individually (from an independent copy of the lists) and tests their inflected, plural, glued, digit-swapped and compatibility-character forms, plus adversarial cases for receipt forgery and PQ stripping; real MCP requests are posted through the whole HTTP stack
The HTTP entrypoint (server_http.py) is written for the container image, where its modules sit side by side; with a pip install, use the stdio server. Bearer auth is off by default and CORS is * in that mode. Anyone who can reach an open deployment can obtain receipts signed with its key, including "ALLOW" receipts for read-only names, and a receipt from an open deployment says nothing about who asked. If you host it, set TRUST_GATE_BEARER_TOKEN and TRUST_GATE_ALLOWED_ORIGINS.
The HTTP entrypoint counts ?via=<label> visits as one log line (label, kind, user-agent family). Only three built-in labels plus those named in TRUST_GATE_CHANNELS (comma separated) are recorded; the host's own access log also records request lines, including client addresses.
DNS-rebinding protection is off and there is no switch for it: on a local HTTP server without bearer auth, any web page you visit can call it. Put it behind a proxy that checks the Host header, or set TRUST_GATE_BEARER_TOKEN.
pip install trust-gate-mcp
trust-gate-mcpAdd [slh] for the native ML-DSA backend (see the table above). From a checkout, pip install -e ".[dev]" then
python -m trust_gate_mcp.
docker build -t trust-gate-mcp .
docker run -p 8081:8081 -v trust-gate-data:/data/oao trust-gate-mcpThe volume mount on /data/oao is required for production: without it the signing key changes on every container restart, receipts signed before a restart still verify from their own certificate, but a pinned kid stops matching new ones. The persistent key_metadata.json holds the signer's kid; the bootstrap step refuses to start if it drifts. The image installs dilithium-py by default; see the table above.
Apache-2.0 (full text in LICENSE). Copyright 2026 Cyber Warrior Network. Built on the open-source OpenAgentOntology primitive.
Rank a CALLER-PROVIDED list of MCP tools by worst-regret if they act (a name-based heuristic that weights a name's first word most). Read-only: it mints no receipt. Cannot auto-discover the inventory -- MCP does not allow that; the caller must pass it in.
| Name | Required | Description | Default |
|---|---|---|---|
| notes | No | ||
| inventory | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations the description carries the full burden, and it does disclose the key traits: read-only with no receipt minted, the name-first-word heuristic nature, and the hard inability to auto-discover inventory. It omits behavior on malformed or empty input (errors, ranking ties), which keeps it short of a 5.
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?
Three sentences, front-loaded with the core action and free of filler. Slightly dense parentheses aside, but every clause adds a distinct fact.
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?
An output schema exists, so return values need no explanation, and the heuristic plus read-only constraint are covered. The unexplained 'notes' parameter and the undocumented structure of inventory entries, against 0% schema coverage, leave it only partially complete for a nontrivial ranking tool.
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?
Schema description coverage is 0% and the description must compensate, but it only characterizes the inventory argument loosely ('list of MCP tools'). The shape of inventory items and the entire 'notes' parameter are left unexplained in both schema and description, leaving a real gap.
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 states a specific verb and resource ('Rank a CALLER-PROVIDED list of MCP tools by worst-regret') and immediately qualifies the scope with the heuristic used. It clearly distinguishes itself from the mint_*/verify_* siblings by stating it mints no receipt, so an agent can route correctly without opening the schema.
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 gives a real usage constraint ('Cannot auto-discover the inventory -- MCP does not allow that; the caller must pass it in') and implies its read-only role versus receipt-minting siblings. However, it never states when to reach for this tool versus check_egress or gate_decision, so selection guidance is only implied.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Egress classification check. Scans a data sample for a finite list of sensitivity markers and classifies as NO_MARKERS_FOUND / INTERNAL / CONFIDENTIAL / RESTRICTED. RESTRICTED sets blocked=true; this tool cannot block anything itself, and NO_MARKERS_FOUND is not clearance to send. Returns classification, retention info, and a tamper-evident receipt.
| Name | Required | Description | Default |
|---|---|---|---|
| provider | Yes | ||
| data_sample | Yes | ||
| destination | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden and does disclose key traits: RESTRICTED drives blocked=true, the tool is non-enforcing, and the result includes retention info and a tamper-evident receipt. It omits whether the sample itself is retained/logged (implied by 'retention info') and any auth or rate-limit requirements.
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?
Three tight sentences, front-loaded with the core action, followed by the exclusivity caveat and return summary. No filler or repetition of the tool name.
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?
An output schema exists, yet the description still summarizes the returns, which is mildly redundant but helpful. Combined with the classification semantics and enforcement boundary, an agent has enough to decide whether to call it; the main gap is undocumented parameter formats.
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?
Schema description coverage is 0% for all three required parameters, so the description must compensate and largely does not. It implies data_sample is what gets scanned but says nothing about the format or meaning of 'provider' and 'destination' (e.g., is provider an email/cloud vendor, is destination a URL?).
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 states a specific verb and resource ('scans a data sample for sensitivity markers and classifies') and names the exact output taxonomy (NO_MARKERS_FOUND / INTERNAL / CONFIDENTIAL / RESTRICTED). This is distinct from siblings like gate_decision or verify_receipt, which handle enforcement and receipts respectively.
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 clearly bounds the tool's role with two strong exclusions: 'this tool cannot block anything itself' and 'NO_MARKERS_FOUND is not clearance to send.' However, it never names the alternative tool (e.g., gate_decision) that does perform blocking, so the agent must infer the routing itself.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Two-phase decision gate with a real verdict. PREVIEW returns ALLOW, DENY or ESCALATE with the risk tier and reasons, plus a preview_id, without acting. COMMIT re-evaluates the same inputs and mints a signed receipt for the verdict. Only ALLOW returns a GRANTED permit; DENY returns DENIED; ESCALATE returns WITHHELD_PENDING_HUMAN and a human must decide outside this tool. Applies only to actions the caller routes through it: it does not observe or block what an agent does. Verb-tier heuristic on the action name and resource, not a proof. Stateless. Optional attestation: triggered_by_type, triggered_by_source, decision_model.
| Name | Required | Description | Default |
|---|---|---|---|
| phase | No | PREVIEW | |
| action | Yes | ||
| context | Yes | ||
| resource | Yes | ||
| preview_id | No | ||
| decision_model | No | ||
| triggered_by_type | No | ||
| triggered_by_source | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden and does so well: it discloses statelessness, that the verdict comes from a verb-tier heuristic on the action name and resource rather than a proof, exactly what each verdict returns, and that ESCALATE requires a human outside this tool. It also warns that the gate does not observe or block un-routed agent actions – a critical scope limit.
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?
Front-loaded with the core concept ('Two-phase decision gate with a real verdict') followed by tightly packed sentences that each carry distinct, load-bearing information. No filler or restatement of the name.
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 stateless, high-complexity tool with an output schema, the description covers the phase model, verdict outcomes, scope limits, and attestation fields adequately; the output schema relieves it of describing return shape. The main gap is `context` semantics and the precise preview_id→COMMIT linkage, but overall it is close to complete.
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?
Schema coverage is 0% across 8 parameters, so the description must compensate. It explains phase (PREVIEW/COMMIT), preview_id's role in the two-phase flow, and that triggered_by_type/source/model are optional attestation, but it never explains the `context` object or what triggered_by_* actually attest, leaving several parameters undocumented in both places.
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 states a specific verb+resource ('two-phase decision gate') and enumerates the exact verdicts (ALLOW/DENY/ESCALATE) and phase semantics (PREVIEW/COMMIT). It is distinguishable from siblings like mint_action_receipt and verify_receipt because it frames this tool as the decision gate that other tools' receipts record/verify.
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 gives clear context: PREVIEW acts without acting, COMMIT re-evaluates and mints a receipt, and it only applies to actions the caller explicitly routes through it. It does not, however, explicitly say when to choose this over mint_action_receipt or verify_receipt, so the alternative-routing guidance is implied rather than stated.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Mint a signed receipt (Ed25519, plus ML-DSA-65 when the post-quantum backend is available) for an arbitrary consequential agent action. Optional attestation: triggered_by_type (human/agent/script), triggered_by_source (api/cli/cron), decision_model (the LLM model used). Attestation values are caller claims: allowlisted to safe characters, signed, not verified. Decisions that contain a gate verdict word are reserved for gate_decision.
| Name | Required | Description | Default |
|---|---|---|---|
| inputs | No | ||
| policy | No | agent action evidence | |
| target | Yes | ||
| agent_id | Yes | ||
| decision | No | ACTION_GOVERNED | |
| operation | Yes | ||
| decision_model | No | ||
| triggered_by_type | No | ||
| triggered_by_source | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the burden, and it does disclose meaningful traits: Ed25519 signing with ML-DSA-65 fallback, attestation values being caller claims (allowlisted, signed, not verified), and the reservation rule. It omits permissions/auth requirements and durability of the emitted receipt.
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?
Front-loaded with the core action and mechanism, then the optional attestation, then the exclusion rule. Dense but each sentence carries information; only the parenthetical backend caveat is slightly incidental.
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?
An output schema exists, so return values need no explanation, and the description covers mechanism, optional fields, and sibling routing. It would be complete if it addressed the overlap with mint_receipt_for_record_change and any auth prerequisites for minting.
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?
Schema coverage is 0% across 9 params, so the description must compensate. It documents three parameters with example values (triggered_by_type human/agent/script, triggered_by_source api/cli/cron, decision_model) and implicitly constrains 'decision' via the gate-verdict rule, but agent_id, operation, target, inputs, policy, and the ACTION_GOVERNED default are left undefined.
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?
States a specific verb+resource ('mint a signed receipt ... for an arbitrary consequential agent action') with the crypto mechanism named, and explicitly carves out the sibling gate_decision for gate-verdict decisions. An agent can distinguish it from verify_receipt and gate_decision without opening a schema.
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?
Explicitly routes gate-verdict-word decisions to gate_decision, which is genuine when-not guidance. However it never mentions mint_receipt_for_record_change, the nearest-scoped sibling, so the agent must infer the boundary between 'arbitrary action' and 'record change'.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Mint a signed receipt (Ed25519, plus ML-DSA-65 when the post-quantum backend is available) for one CRM record change. Old/new values are carried as SHA-256 hashes. Works with any CRM (Relaticle, hosted CRMs, custom).
| Name | Required | Description | Default |
|---|---|---|---|
| field | Yes | ||
| policy | No | per-decision CRM change evidence | |
| tenant | No | ||
| new_value | Yes | ||
| old_value | Yes | ||
| record_id | Yes | ||
| object_type | Yes | ||
| changed_by_agent | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden, and it does add real behavioral context: signature algorithms (Ed25519, conditional ML-DSA-65), SHA-256 hashing of values, and CRM-agnostic support. It omits important traits for a mutation-style tool, such as whether receipts are persisted, idempotency, required permissions, or failure behavior when the post-quantum backend is unavailable.
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?
Three tight sentences, front-loaded with the core purpose. Some budget is spent on crypto-backend detail that matters less than the undocumented parameters, but there is no filler.
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?
An output schema exists so return values need no explanation, but for an 8-parameter, annotation-free tool the description leaves too much unsaid. No parameter guidance, no usage routing, and no operational constraints make it insufficient for correct invocation.
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?
Schema description coverage is 0% across 8 parameters (6 required), so the description must compensate and does not. It never explains record_id, object_type, field, tenant, policy, or changed_by_agent, and its claim that old/new values are 'carried as SHA-256 hashes' ambiguously implies hashing behavior without clarifying whether callers pass raw or pre-hashed values.
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?
States a specific verb (mint) and resource (signed receipt for one CRM record change) plus scope, so an agent knows what it produces. However, it does not distinguish itself from the sibling mint_action_receipt, leaving the agent to guess which receipt tool applies to which scenario.
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?
No when-to-use guidance, no prerequisites, and no routing to or from alternatives like mint_action_receipt or verify_receipt, despite those siblings existing. The agent is left to infer the place of this tool in the broader workflow.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Vendor exit readiness drill. Checks local signing key, local model access (Ollama). Returns step-by-step results and a tamper-evident receipt. Informational; it signs one receipt, which creates the signing key on a host that has none yet.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full behavioral burden, and it does disclose a genuinely non-obvious side effect: despite being 'Informational', it signs one receipt, which creates a signing key on a host that has none yet. That is exactly the kind of mutation disclosure an agent needs before invoking. It stops short of describing auth requirements, rate limits, or failure modes.
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?
Three short sentences, front-loaded with what the drill is, then what it checks, then the return shape and the caveat. Every sentence carries information and none is redundant.
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?
An output schema exists, so the description need not explain the step-by-step results or receipt structure, and it correctly does not. It covers purpose, checks performed, and the key side effect; only invocation prerequisites (permissions, environment assumptions) are left implicit.
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 tool takes zero parameters and the schema is empty at 100% coverage, so there is nothing for the description to disambiguate. The baseline for a parameterless tool is 4.
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 names a specific verb+resource ('Vendor exit readiness drill') and enumerates exactly what it inspects: the local signing key and local model access via Ollama. It is clearly not a receipt-minting or egress-checking tool, so an agent can separate it from most siblings, though it never explicitly contrasts itself with check_egress or audit_my_agent_inventory.
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?
The word 'Informational' implies this is a diagnostic to run when validating vendor-exit readiness, and the check list hints at the preconditions it evaluates. However, there is no explicit when-to-use, when-not-to-use, or named alternative among the six sibling tools, so usage must be inferred.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Verify a Trust Gate receipt from the certificate alone (offline). ok=true means unchanged since signing, signed, kid consistent with the embedded key, and (require_pq default on) at least one post-quantum leg verified; unsigned receipts fail. It does not prove who signed: pass expected_kid to pin the signer's Ed25519 key; a pin counts only when that key's own signature verifies (see signer_pinned). The post-quantum keys in a receipt are not covered by the kid.
| Name | Required | Description | Default |
|---|---|---|---|
| receipt | Yes | ||
| require_pq | No | ||
| expected_kid | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description fully carries the behavioral burden. It explains what ok=true means, that unsigned receipts fail, that it does not prove signer identity, how expected_kid pinning works, and that post-quantum keys are not covered by the kid.
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?
The description is front-loaded with the core verification contract, then adds the important signer-pinning caveat and PQ-key caveat. Every sentence adds necessary semantic information without filler.
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?
Given the cryptographic complexity, missing annotations, and 0% schema description coverage, the description covers the critical behavioral guarantees and limitations. Return values are additionally covered by the output schema.
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?
Schema description coverage is 0%, so the description must compensate. It explains require_pq's effective default and expected_kid's pinning semantics in detail, though the receipt parameter itself is left largely opaque as a Trust Gate receipt object.
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 states a specific verb and resource: verify a Trust Gate receipt. It also names the exact verification mode, offline from the certificate alone, which clearly distinguishes it from the minting siblings.
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 gives clear context for use and conditionally describes when to supply expected_kid to pin the signer. It does not explicitly compare against sibling tools or state when not to use this tool, so it falls short of the top level.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.3.0verify_receipt1 field changedInput schema / properties / expected_kidAdded value: +{
+ "anyOf": [
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "title": "Expected Kid"
+}v0.2.1audit_my_agent_inventorycheck_egressgate_decisionmint_action_receiptmint_receipt_for_record_changerun_exit_drillverify_receiptScored across 7 tools
Receipt minting appears in mint_receipt_for_record_change, mint_action_receipt, gate_decision, and implicitly check_egress/run_exit_drill, but detailed descriptions separate CRM record changes from arbitrary actions and gate verdicts. A caller might still momentarily confuse mint_action_receipt with gate_decision when an action involves a gate verdict word, though the reserved-word note mitigates this.
Almost all names use snake_case with a leading verb (verify_, mint_, audit_, check_, run_), and the pattern is easy to scan. Minor deviations: gate_decision reads more like a noun phrase than a verb_noun action, and audit_my_agent_inventory adds 'my' and is longer than the others.
Seven tools is well within the ideal 3-15 range for a narrowly scoped trust/receipt server. Each tool covers a distinct operation — verify, mint two receipt types, gate, egress classify, audit inventory, exit drill — so none feels redundant or purely ornamental.
Core receipt lifecycle is present: mint signed receipts, verify offline, gate decisions, egress classification, inventory audit, and exit drill. Missing a few ancillary lifecycle operations such as explicit key management/rotation, receipt listing/retrieval, or a post-escalation human decision path, but these are either outside scope or intentionally external.
Independent effect verification and signed receipts for consequential AI agent actions.
Did it really happen? Independent signed receipts for AI agents: page, domain, email, heartbeat.
Issue signed receipts for AI agent actions; verify any receipt offline - free, no account.
Cryptographically anchored evidence for agents: verified run receipts, proof-gated settlement.
