Documentation Index
Fetch the complete documentation index at: /llms.txt
Use this file to discover all available pages before exploring further.

identity

GET
Get the current caller's identity, role, and KB clearance
GET
Get current organization
PUT
Update current organization
GET
Get the org's warm-transfer settings (bridge webhook URL)
GET
List the org's warm-transfer bridge numbers
POST
Import a Twilio number as a warm-transfer bridge number
DEL
Delete a bridge number (clears its Twilio webhook, then drops the row)
POST
Re-apply the current /twilio/{org_id}/bridge webhook URL to all bridge numbers
GET
List organization members
POST
Resolve a channel source to an end user (graph-backed)
GET
List end users for the organization (optionally filter by org_external_end_user_id)
GET
Get an end user with their contacts
PUT
Update an end user (auto-merges on org_external_end_user_id conflict)
DEL
Delete an end user and all associated data
POST
Attach a verified contact to an end user
GET
List identity review requests for the organization
GET
Get a single identity review request
POST
Approve a MERGE review request and execute the merge
POST
Approve an UPDATE review request and add the proposed contact
POST
Reject a review request
GET
List API keys for the current organization
POST
Create an API key. The plain-text key is returned exactly once.
DEL
Revoke an API key

sessions

GET
Fetch the canonical analysis input for a completed session

retention-policies

GET
List retention policies
POST
Create a retention policy
PUT
Update a retention policy
DEL
Delete a retention policy

memory

GET
Get all facts and entities for a user
DEL
Delete all memories for a user (GDPR)

knowledge-base

GET
Get available KB configuration options
GET
List knowledge bases
POST
Create a knowledge base
GET
Get a knowledge base
PUT
Update a knowledge base
DEL
Delete a knowledge base
GET
Get aggregate ingestion status for a knowledge base
GET
List documents in a knowledge base
GET
Get a document
DEL
Delete a document
PATCH
Update a document's title and/or metadata
GET
Get a document's sensitivity classification
PATCH
Manually correct a document's sensitivity classification
POST
Approve a document's pending classification
GET
List documents awaiting classification review
POST
Re-run sensitivity classification across a KB's documents
POST
Ingest documents (text, files, or mixed)
GET
List ingestion jobs for a knowledge base
GET
Get ingestion job progress
GET
List documents in an ingestion job
GET
List ingestion job logs grouped by document
POST
Retry a failed ingestion job
POST
Semantic search across knowledge bases
GET
List all versions of a document
GET
Get a version by UUID, version number, or alias
DEL
Delete a specific version (blocked if active)
POST
Promote a version to active
POST
Release a quarantined version (reviewed false-positive)
POST
Confirm a quarantined version (true-positive; stays quarantined)

knowledge-base-evals

GET
List evaluation suites
POST
Create an evaluation suite
GET
Get an evaluation suite with its cases
DEL
Delete an evaluation suite
PATCH
Update an evaluation suite
POST
Add a case to an evaluation suite
DEL
Delete an evaluation case
PATCH
Update an evaluation case
GET
List evaluation runs for a suite
POST
Trigger a new evaluation run against a suite
GET
List evaluation runs across the org (including runs whose suite was deleted)
GET
Get an evaluation run
GET
List per-case results for a run

knowledge-base-connectors

GET
List Sources
POST
Create Source
GET
Get Source
DEL
Delete Source
GET
List Roots
POST
Add Root
DEL
Delete Root
GET
Drive Picker Config
GET
Browse Pages
POST
Trigger Sync
GET
Get Sync Status
PATCH
Update Source Schedule

knowledge-base-governance

PATCH
Update a version's lifecycle governance status / expiry
GET
List a version's governance lifecycle history (newest first)
GET
Get the org's KB audit retention policy (creates defaults if unset)
PUT
Update the org's KB audit retention policy / legal hold

knowledge-base-audit

GET
Query governed KB retrieval-audit logs (incl. misses/blocks/errors)
GET
Get one retrieval-audit log with its scopes and chunk refs
POST
Request an async encrypted KB audit export
GET
Get an audit export job's status + per-file download tokens
GET
Server-mediated, decrypted download of one export file

knowledge-base-acl

GET
Get the org's effective KB ACL config (role→clearance map + mode)
PUT
Update the org's KB role→clearance map and/or classification mode
POST
Enqueue the KB ACL enable preflight (verifies coverage + config)
GET
Get the org's KB ACL rollout state + last preflight report
POST
Enable KB ACL (requires a passing preflight → ready_to_enable)
POST
Disable KB ACL (returns retrieval to byte-identical pre-ENG-640)

model-router

GET
List available models with pricing and capabilities
GET
List selectable models plus per-role runtime model defaults
GET
List model router configs
POST
Create a model router config
GET
Get a model router config
PUT
Update a model router config
DEL
Delete a model router config
GET
List model fallback events for auditing
POST
Invoke a model with routing
POST
Stream model responses via SSE

notifications

GET
Discover channels the bot can post to (cached 5 min)
GET
List configured notification destinations
POST
Create a destination (verified on save via a test post)
DEL
Delete a notification destination
GET
List event-type routing rules
PUT
Create or update an event_type -> destination route
DEL
Delete a routing rule
GET
Get org notification settings (default destination + timezone)
PUT
Set the default destination and/or org timezone
GET
Browse the notification catalog merged with this org's config
PUT
Enable/disable a type and set its severity override / threshold config
POST
One-click enable the recommended starter set
POST
Send a test notification through the real dispatch path (notify())
GET
Browse the delivery-decision audit log (why a notification was/wasn't sent)
GET
Get the effective template for a type (default or override) + variables manifest
PUT
Create/update the per-org template override (validates + increments version)
DEL
Delete the per-org override (revert to the platform default)
POST
Strict-validate candidate slots (per-slot errors; valid=false is a 200)
POST
Render a preview (text + blocks) of candidate-or-effective slots

integrations

GET
List the integrations this platform supports
GET
List the calling org's connections
POST
Start the OAuth flow for a new connection
GET
Get a connection
DEL
Disconnect (hard-delete the row + drop vendor-side connection)
PATCH
Rename a connection (slug stays immutable)
POST
Connect a customer-supplied MCP server (static creds or OAuth)
POST
Connect the Email integration (born active; presence marker for the send_email tool)
POST
Pre-check a custom MCP server's OAuth/DCR support
POST
Finalize a connection after the OAuth popup closes
POST
Manually re-sync the tool list for this connection
PATCH
Enable or disable all tools on a connection
PATCH
Edit author-editable metadata on a single tool (response_field_map, is_enabled, input_defaults — defaults + pinned)
POST
Dashboard test surface: invoke a tool by slug (VENDOR_API)
POST
Dashboard test surface: invoke an MCP tool by slug (FEATHER_MCP_CLIENT)
POST
Dashboard test surface: lease a provider token (FEATHER_API)
GET
Recent webhook events for the org (testing/observability)

tools

GET
List API tools
POST
Create an API tool
GET
Get an API tool by ID
PUT
Update an API tool
DEL
Delete an API tool
POST
Execute an API tool
POST
Preview a revision's RAW response + narrowed field map (no persist)
POST
Enable an API tool
POST
Disable an API tool
GET
List tool revisions
POST
Create a tool revision
GET
Get a tool revision
PUT
Update a tool revision
DEL
Delete a tool revision
POST
Set the active tool revision

mocks

GET
List mocks
POST
Create a mock
GET
Get a mock
PUT
Update a mock
DEL
Delete a mock
POST
Enable a mock
POST
Disable a mock
GET
List scenarios on a mock (priority ASC)
PUT
Atomically replace all scenarios on a mock
POST
Create a scenario for a mock
POST
Bulk update scenario priorities
GET
Get a scenario
PUT
Update a scenario
DEL
Delete a scenario
GET
Paginated, filterable list of dispatcher requests for a mock
GET
Get a single dispatcher request log entry (verbose)

hitl

GET
Get the org's HITL approval-UX lines (resolved)
PUT
Set/clear the org's HITL approval-UX lines
GET
List approval requests
GET
Get an approval request
POST
Decide an approval request (approve / deny)
GET
List handoffs
GET
Get a handoff
POST
Relay an operator message on a handed-off conversation
POST
Resolve a handoff and close its conversation
GET
List failed warm-transfer attempts

personas

GET
List personas
POST
Create a persona
GET
Get a persona
PUT
Update a persona
DEL
Delete a persona

runtime v2

GET
List v2 conversations
POST
Get or create a v2 conversation
GET
List a v2 conversation's transcript
POST
Run one turn of a v2 conversation
GET
Get a v2 conversation with recent turns
DEL
Delete a v2 conversation and purge its data + blobs
POST
Close a v2 conversation
POST
End a v2 conversation through the close pipeline (admin)
POST
Mint a fresh test-type v2 conversation bound to an agent or team
GET
Poll a v2 conversation for its status and the message tail since a seq
GET
Get the customer-readable trace for a v2 conversation
GET
Get the customer-readable trace for a single v2 turn
GET
List the per-turn engineering-waterfall summaries for a v2 conversation
GET
Get the reconstructed engineering waterfall for a single v2 turn
POST
Re-run all current evaluator bindings against a v2 conversation
POST
Run one v2 conversation turn with an SSE response

agents

GET
List agents
POST
Create an agent
GET
List declared context variables across an agent's bound workflows
GET
Get an agent
PUT
Update an agent
DEL
Delete an agent
GET
Get per-agent eval configuration
PUT
Set per-agent eval configuration
GET
List the platform tools an agent can enable
GET
Get an agent's enabled platform tools
PUT
Set an agent's enabled platform tools
GET
List agent revisions
POST
Create an agent revision
GET
Get an agent revision
PUT
Update an agent revision
DEL
Delete an agent revision
POST
Set active agent revision
GET
Latest release-check verdict for a candidate revision
POST
Dispatch the agent's release simulation suite against a candidate revision
GET
Get an agent revision's tool-input defaults (enriched view)
PUT
Replace an agent revision's tool-input defaults
GET
List workflows attached to agent revision
POST
Attach a workflow to agent revision
PUT
Update workflow link on agent revision
DEL
Detach workflow from agent revision

policies

GET
List policies
POST
Create a policy
GET
Get the org-wide policy enforcement switch
PUT
Enable/disable all policy enforcement for the org (kill switch)
GET
Get the org-level llm_judge model default
PUT
Set/clear the org-level llm_judge model default
GET
List policy evaluation audit rows (what enforcement happened, newest-first)
GET
Describe the expression-check DSL (paths per point, predicates, operators)
GET
Describe the policy authoring cascade + per-transport runtime resolution
GET
Resolve a tool's input/output schema for policy authoring
GET
Get a policy
PUT
Update a policy's identity fields
DEL
Delete a policy
GET
List policy revisions
POST
Create a policy revision
GET
Get a policy revision
DEL
Delete a policy revision
POST
Set active policy revision
GET
Get an assistant revision's attached policies
PUT
Attach/detach policies on an assistant revision
DEL
Detach all policies from an assistant revision
GET
Describe an assistant's policy enforcement plan (per point, per transport)

teams

GET
List teams
POST
Create a team
GET
Get a team
PUT
Update a team
DEL
Delete a team
GET
List team revisions
POST
Create a team revision
GET
Get a team revision
POST
Set active team revision
GET
List members on a team revision
POST
Add a member to a team revision
PUT
Update a member on a team revision
DEL
Remove a member from a team revision

scenarios

GET
List scenarios
POST
Create a scenario
POST
Generate an unsaved scenario draft from a brief
POST
Generate an unsaved scenario draft from a captured session
GET
Get a scenario
PUT
Update a scenario
DEL
Delete a scenario

sim_suites

GET
List simulation suites
POST
Create a simulation suite
GET
Get a simulation suite
PUT
Update a simulation suite
DEL
Delete a simulation suite
POST
Clone a simulation suite
POST
Add an item to a simulation suite
PUT
Update a simulation suite item
DEL
Remove a simulation suite item
GET
List simulation suite runs for a suite
POST
Dispatch a simulation suite run (fan-out per item)
GET
Get a simulation suite run with its child runs
POST
Cancel every non-terminal child of a suite run
POST
Re-dispatch every failed child of a suite run

simulation_runs

GET
List simulation runs
POST
Dispatch a simulation run
GET
Get a simulation run
GET
Get the customer-readable trace for a simulation run
GET
Get the customer-readable trace for one turn of a simulation run

evals

GET
List evaluators
POST
Create an evaluator
GET
Get an evaluator
PATCH
Update an evaluator
POST
Archive an evaluator
GET
List evaluator bindings
POST
Create an evaluator binding
GET
Get an evaluator binding
DEL
Delete an evaluator binding
PATCH
Update an evaluator binding
GET
List eval runs
GET
Get an eval run (optionally with embedded evidence)
DEL
Delete a terminal eval run
POST
Retry a failed or errored eval run
GET
Get an eval score
GET
List deterministic signals available to evaluators
POST
(dev only) Fabricate failing eval runs and fire the real eval.failed emitter
POST
Re-run all current evaluator bindings against a session

library

GET
Browse the persona library
POST
Add a persona to the global library (Feather staff only)
GET
Get a library persona
PUT
Update a library persona (Feather staff only)
DEL
Delete a library persona (Feather staff only)
POST
Install a library persona into the caller's workspace
GET
Get the voice preset attached to a library persona
POST
Attach a voice preset to a library persona (Feather staff only)
DEL
Detach the voice preset from a library persona (Feather staff only)
PATCH
Update the voice preset on a library persona (Feather staff only)

privacy

GET
Get privacy config for organization
POST
Create privacy config for organization
DEL
Delete privacy config (reverts to defaults)
PATCH
Update privacy config for organization
GET
List PII scrub audit events
GET
Aggregate PII scrub stats

communication

GET
List org phone numbers
POST
Register an org phone number
GET
Get a phone number by id (org-scoped)
DEL
Delete a phone number
PATCH
Update a phone number
GET
List channel bindings for a phone number
POST
Bind an agent or team to a channel on a phone number
DEL
Remove a channel binding
PUT
Replace the full binding set for a channel on a phone number
GET
Get the org's compliance settings (creates default on first read)
PUT
Update the org's allowed-countries list

sms

POST
Send an outbound SMS (queued; delivered asynchronously)
GET
Get the org's SMS settings
PUT
Update the org's SMS settings (partial)

email

POST
Begin connecting a mailbox via the hosted-auth flow
GET
List connected mailboxes for the caller's organization
DEL
Revoke the mailbox grant and retire the local row
PATCH
Update mutable mailbox settings (e.g., default signature)
GET
Read an email thread by Nylas provider_thread_id
POST
Ingest a canonical EmailInboundEvent for the given thread. Bypasses the vendor webhook; used by sandbox / replay / direct integrators.
GET
Delivery event audit trail for a single email message

email-drafts

GET
List drafts for the caller's organization
POST
Author a new outbound email draft
GET
Read one draft, including its ENG-420 audit snapshot when present
DEL
Soft-discard a draft (forbidden once dispatch has started)
PATCH
Edit a draft (only allowed while status='draft')
POST
Approve a draft and dispatch it. The caller's OrgMember is recorded as the approver; the SendOrchestrator runs immediately.

voice

GET
List voice pipeline configs for the org
POST
Create a voice pipeline config for an agent or persona (1:1)
GET
Get a voice pipeline config by id
DEL
Delete a voice pipeline config
PATCH
Partial update of a voice pipeline config
GET
Get the voice pipeline config for an agent
GET
Get the team-level voice pipeline config for a team
GET
Get the voice pipeline config for a persona
GET
List Voices
GET
Get Voice
GET
List Tts Models
GET
List Stt Models
GET
Get Recording Playback Url
GET
Get Warm Transfer Detail
POST
Outbound
POST
Web Call
POST
Ingest a voice call (ASR + PII scrub + optional audio retention)
GET
Get a voice call's scrubbed transcript and audio URL

analytics

GET
Tool invocations bucketed over time
GET
Tool invocations grouped by tool_type
GET
Top-N tools by invocation count
GET
Tool invocations grouped by agent_id
GET
Distinct conversations per tool
GET
Latency percentiles bucketed over time
GET
Top-N slowest tools by p95
GET
Latency histogram across fixed buckets
GET
Latency percentiles grouped by provider
GET
Per-retry-count breakdown with avg/p95 latency
GET
Success rate (%) bucketed over time
GET
Failed-invocation count grouped by error_category
GET
Failed-invocation count grouped by HTTP status_code
GET
Per-tool success rate ranked worst-first
GET
Error count over time, split by error_category
GET
Current vs previous window error-rate comparison
GET
Per-tool timeout rate and latency tail
GET
Combined tool analytics dashboard envelope
GET
New-session counts bucketed over time, split by channel
GET
Current distribution of sessions across statuses
GET
Resolution-time percentiles by channel and priority
GET
Time-to-first-response percentiles
GET
Per-channel comparison of session volume and quality
GET
Per-bucket escalation rate timeseries
GET
Histogram of message count per session
GET
Turn-depth summary (avg / single-turn rate / quantiles)
GET
Per-KB adoption rate and document counts
GET
Document-ingestion counts bucketed over time
GET
Document processing-time percentiles bucketed over time
GET
Failure-recovery analytics: failed / archived / cleaned_up / recovered
GET
Search-call volume and latency percentiles bucketed over time
GET
Per-agent retrieval breakdown: volume, hit-rate, and avg latency
GET
Top-N most-retrieved documents in a window
GET
Documents currently active but unretrieved (or stale beyond N days)
GET
Current-state KB inventory shape
GET
Per-source latest sync state and time-since-last-event
GET
One-shot envelope of every knowledge analytics pipe
GET
Memory episode counts bucketed over time
GET
Memory coverage: users-with-memory / active-end-users (clamped 100%)
GET
Memory recall volume bucketed over time with hit/miss + latency percentiles
GET
Per-vendor recall summary (zep vs mem0)
GET
Per-user engagement quantiles (episodes/messages/recalls)
GET
One-shot envelope of every memory analytics pipe
GET
Single-row dashboard summary: requests, match/error rates, latency
GET
Mock request volume bucketed over time
GET
Mock latency percentiles bucketed over time
GET
Mock scenario match rate bucketed over time
GET
Top-N mock scenarios by request count
GET
One-shot envelope of every mock analytics pipe
GET
Single-row summary: invocations, failures, success rate, latency (vs prior window)
GET
Integration tool-call volume bucketed over time
GET
Effective success rate (%) bucketed over time
GET
Integration tool-call latency percentiles bucketed over time
GET
Volume + reliability + p95 latency grouped by integration_key
GET
Volume + reliability + p95 latency grouped by mode (vendor_api/feather_mcp_client)
GET
Failures grouped by error_category (vendor_error / exception:<type>)
GET
One-shot envelope of every integration analytics pipe
GET
Single-row connection-health summary: events, failures, success rate, latency
GET
Connection health grouped by integration_key (events, success rate, p95 latency)
GET
Connection lifecycle event volume (events/successes/failures) over time
GET
Connection lifecycle effective success rate (%) over time
GET
Connection lifecycle failures grouped by event_type + error_category
GET
Per-policy violation rate + action counts within a window
GET
Action counts split into taken (enforced) vs would-be (monitor-mode shadow)
GET
Evaluation + violation volume bucketed over time (enforced vs shadow)
GET
Enforce-after-measure: per monitor-mode policy, would-be impact + TTFT-delta
GET
One-shot envelope of every policy analytics pipe
GET
Whether an enforcing agent_response policy is attached (TTFT disclaimer)

knowledge-gap-analytics

GET
Headline knowledge-gap KPIs for the org over a date window
GET
Paginated list of knowledge-gap clusters with filters and sort
GET
Detail view of a single knowledge-gap cluster
GET
Resolve conversation turns referenced by knowledge-gap evidence
GET
Per-session analysis result, signals, gap events, and linked clusters
GET
Per-document retrieval / citation / implication counts with failure rate
PATCH
Assign or clear the owner of a knowledge-gap cluster
POST
Transition a cluster through the lifecycle state machine
GET
Audit-trail of lifecycle transitions for a cluster

extraction-schemas

GET
List extraction schema versions for an agent
POST
Create a new extraction schema version for an agent
GET
Get one extraction schema version
PATCH
Update extraction schema (rename, activate, archive)

extractions

GET
Get latest structured extraction for a session
POST
Re-run extraction for a closed session

mcp-chat

GET
List the MCP tools available to the chat surface
POST
Stream an LLM chat turn that uses the MCP server as its toolbelt

ask-feather

GET
Get Workflows
GET
List Sessions
POST
Create Session
GET
Get Session Detail
DEL
Archive Session
POST
Send Message
POST
Approve Message

webhooks

POST
Create a hosted Svix App Portal link for the caller's org
GET
List supported customer webhook event types

workflows

GET
List workflows
POST
Create a workflow
GET
Get a workflow
PUT
Update a workflow
DEL
Delete a workflow
POST
Bulk delete workflows
GET
List workflow revisions
POST
Create a workflow revision
GET
Get a workflow revision
PUT
Update a workflow revision
DEL
Delete a workflow revision
GET
Diff two workflow revisions
GET
Get a workflow revision's compiled graph
PUT
Validate and manually replace a workflow revision's compiled graph
POST
Validate a workflow graph without saving it
GET
Get a workflow revision's tool-input defaults (enriched view)
PUT
Replace a workflow revision's tool-input defaults (draft-only)
POST
Set the active workflow revision
POST
Compile a workflow revision (async)
POST
Fork a workflow revision into a draft and recompile it
POST
Cancel an in-flight workflow compile (best-effort)

workflow-playground

POST
Create Session
POST
Send Message
POST
Resume Approval
GET
Get Session Info

Get API Access

Feather home page

Feather

Get API Access
Get API Access

List declared context variables across an agent's bound workflows

Aggregate the union of context_variables declared across the workflows bound to the agent’s active revision.

Used by the evaluator-create UI to drive a typo-safe dropdown for the agent.variable_extracted deterministic-eval signal. The service owns the union + dedup (agent-own wins) and returns typed ContextVarSpec — this handler just maps each spec to the response shape.

GET

agents

{agent_id}

variables

List declared context variables across an agent's bound workflows

curl --request GET \
  --url https://api-sandbox.featherhq.com/v1/agents/{agent_id}/variables \
  --header 'x-api-key: <api-key>'

{
  "variables": [
    {
      "name": "<string>",
      "type": "<string>",
      "source": "<string>",
      "description": "<string>",
      "enum_values": [
        "<string>"
      ]
    }
  ]
}

Authorizations

x-api-key

string

header

required

Path Parameters

agent_id

string<uuid>

required

Response

Successful Response

variables

AgentVariableResponse · object[]

required

Show child attributes

Was this page helpful?

Suggest edits Raise issue

Get an agent

⌘I

Powered byThis documentation is built and hosted on Mintlify, a developer documentation platform

List declared context variables across an agent's bound workflows

curl --request GET \
  --url https://api-sandbox.featherhq.com/v1/agents/{agent_id}/variables \
  --header 'x-api-key: <api-key>'

{
  "variables": [
    {
      "name": "<string>",
      "type": "<string>",
      "source": "<string>",
      "description": "<string>",
      "enum_values": [
        "<string>"
      ]
    }
  ]
}

Assistant

Responses are generated using AI and may contain mistakes.