MetatextAI MCP integration for AI agents.
Connect AI agents to MetatextAI through 4 structured actions, including list policies, create policy, and evaluate. Review authentication, inputs, workflows…
- ACTION CONTRACTS
- 4
- AUTHENTICATION
- API key
- SOURCE STATUS
- Provider-backed
CATALOG SOURCE REVIEWED AUGUST 23, 2026 / ACTION NAMES AND SCHEMAS DERIVED FROM WORKING MACHINES PROVIDER SOURCE
CAPABILITY PROFILE
What the MetatextAI integration exposes
The MetatextAI Agent App exposes 4 provider-backed actions for ai and security work. Its current contract lets an authorized agent list all guardrail policies configured for the connected metatextai application, create one guardrail policy for the connected metatextai application, and evaluate one chat transcript against the connected metatextai application's configured guardrails. These operations are called through Working Machines as typed capabilities rather than through browser navigation or copied UI steps.
A connection uses API key. Before execution, the agent can inspect the selected action, its required fields, declared scopes, and expected output contract. Provider credentials remain inside the Working Machines runtime; the calling agent receives the capability and its structured result, not the underlying secret.
For reliable operation, start with the narrowest action that satisfies the task, resolve stable provider identifiers before changing state, and validate the returned object or status after execution. Availability still depends on the connected MetatextAI account, granted provider permissions, workspace policy, region, plan, and upstream API behavior.
VERIFIED ACTION SAMPLE
Real MetatextAI capabilities.
Showing 4 of 4 actions. Risk labels are conservative signals based on operation names, not substitutes for provider documentation or runtime policy.
list_policiesREADList all guardrail policies configured for the connected MetatextAI application.
create_policyWRITECreate one guardrail policy for the connected MetatextAI application.
INPUTS: name / type / target / rule
evaluateREVIEWEvaluate one chat transcript against the connected MetatextAI application's configured guardrails.
INPUTS: messages / policyIds / failFast / correctionEnabled / overrideResponse
run_test_scanWRITERun a MetatextAI red-team test scan for the connected application with the selected probes.
INPUTS: probes
INPUT CONTRACTS
Know what the action needs before it runs.
create_policy
nameREQUIRED- Policy name shown in MetatextAI.
typeREQUIRED- Policy type identifier accepted by MetatextAI.
targetOPTIONAL- Application message targets checked by this policy.
ruleOPTIONAL- Policy rule payload accepted by MetatextAI. Keep nested fields flexible because rule shape varies by policy type.
evaluate
messagesREQUIRED- Messages checked by MetatextAI.
policyIdsOPTIONAL- Optional policy identifiers to evaluate explicitly.
failFastOPTIONAL- Whether MetatextAI should stop at the first violation.
correctionEnabledOPTIONAL- Whether MetatextAI should return a corrected output when a violation is found.
overrideResponseOPTIONAL- Optional fixed response string returned instead of the blocked model output.
run_test_scan
probesOPTIONAL- Probe identifiers to run in the red-team scan.
PROVIDER-SPECIFIC WORKFLOWS
Jobs this Agent App can support
Inspect List Policies
List all guardrail policies configured for the connected MetatextAI application. Use this as a bounded discovery step, retain the returned identifier, and avoid expanding the read beyond the task's stated scope.
list_policiesControl Create Policy
Create one guardrail policy for the connected MetatextAI application. Resolve the target first, present material changes for confirmation, and make retries idempotent where the provider supports it.
create_policyVerify with Create Policy
Create one guardrail policy for the connected MetatextAI application. Compare the returned provider state with the intended outcome and preserve stable IDs or canonical links in the run record.
create_policySAFETY BOUNDARY
Operate MetatextAI with explicit limits
- Authorize MetatextAI with API key and grant only the provider access required by the selected actions.
- 2 actions are change-capable by name. Confirm the target identity and material parameters before allowing a write.
- Do not infer permission from catalog visibility. Workspace policy, connection identity, and upstream authorization still govern execution.
- No provider scope string is declared on the sampled actions. Verify the connected account's actual permissions in MetatextAI rather than assuming unrestricted access.
- After a call, inspect the structured result and execution record before reporting that the MetatextAI task completed successfully.
CONNECTION MODEL
API key
MetatextAI auth token used with the Authorization: Bearer <token> header for guard-api requests. Create or copy it from the MetatextAI console API/auth token settings documented in the official API reference: https://docs.metatext.ai/introduction.
Working Machines stores provider credentials behind the execution boundary. An agent can use an authorized connection identity, but catalog discovery alone does not reveal OAuth tokens, API keys, or provider secrets.
- READ SIGNALS
- 1
- WRITE SIGNALS
- 2
- HIGH IMPACT
- 0
- REVIEW SIGNALS
- 1
EVIDENCE AND AVAILABILITY
Provider reference
Action names, input fields, authentication types, and counts on this page are generated from the Working Machines provider catalog. Provider behavior, quotas, object semantics, account eligibility, and regional availability remain governed by MetatextAI.