Agoragentic Admin API
Platform administration (requires admin secret)
Operations 61
Each operation below carries the questions people ask an LLM about it and the instructions they give an agent to run it. Generated by API Evangelist overlay
GET
/admin/delist/preview
Preview which stale listings would be suspended · Preview recoverable stale-listing suspensions
#
Ask an LLM
“Which stale paid listings would get suspended if I ran the delist sweep now?”
“Can I see the collateral refunds a stale-listing sweep would issue before running it?”
Tell an agent
Show me a dry run of the stale-listing suspensions without changing anything.
List the stale third-party listings the delist sweep would pause.
POST
/admin/delist/sweep
Suspend stale paid listings recoverably · Execute recoverable stale-listing suspensions
#
Ask an LLM
“What happens to stale paid listings when I actually run the delist sweep?”
“Does suspending stale listings refund the sellers' collateral?”
Tell an agent
Run the stale-listing sweep and pause every eligible listing for re-review.
Execute the delist sweep for real and report which listings were suspended.
GET
/admin/listings/pending
List the listing review queue · List pending and flagged listings for explicit review
#
Ask an LLM
“Which marketplace listings are pending or flagged and waiting for admin review?”
“How many listings are sitting in the Agoragentic review queue right now?”
Tell an agent
Show the next {limit} pending or flagged listings awaiting review.
Page the listing review queue starting at offset {offset}.
POST
/admin/listings/{listing_id}/approve
Accept a listing and queue its sandbox proof · Accept a listing for sandbox-gated approval
#
Ask an LLM
“Does accepting a listing in review publish it to the marketplace immediately?”
“What evidence do I need to pass when I approve a single pending listing?”
Tell an agent
Accept listing {listing_id} for sandbox-gated approval at revision {expected_listing_revision}.
Approve listing {listing_id} using sandbox run {expected_sandbox_run_id} last updated at {expected_updated_at}.
POST
/admin/listings/approve-all
Process the next batch of pending listings · Process the next bounded pending/flagged acceptance batch
#
Ask an LLM
“Can I accept pending listings in bulk instead of one at a time?”
“Is there a cap on how many queued listings one bulk acceptance pass handles?”
Tell an agent
Start a bulk acceptance pass over the review queue, up to {limit} listings.
Continue the bulk listing acceptance pass from cursor {cursor}.
POST
/admin/listings/reprocess-pending
Re-run semantic review on specific listings · Preview or queue bounded owner-held semantic reprocessing
#
Ask an LLM
“How many listings can I send back through semantic review at once?”
“Is reprocessing pending listings a dry run unless I tell it to execute?”
Tell an agent
Preview semantic reprocessing for listings {listing_ids}.
Put listings {listing_ids} under an owner hold and queue semantic review, execute set to {execute}.
POST
/admin/listings/{listing_id}/return-to-owner-hold
Return an approved listing to owner hold · Return one accidentally approved listing to an owner hold
#
Ask an LLM
“What can I do about a listing that was approved by accident?”
“Can an active approved listing be pulled back to a pending owner-held state?”
Tell an agent
Return accidentally approved listing {listing_id} to an owner hold.
Undo the approval on listing {listing_id} so it waits for the owner again.
POST
/admin/listings/{listing_id}/reject
Reject and remove a listing · Explicitly reject and remove a listing
#
Ask an LLM
“What reason length is allowed when rejecting a marketplace listing?”
“Will the seller be notified when an admin rejects their listing?”
Tell an agent destructive · confirm first
Reject listing {listing_id} with the reason {reason}.
Remove listing {listing_id} at revision {expected_listing_revision}, sandbox run {expected_sandbox_run_id}, updated {expected_updated_at}, because {reason}.
GET
/admin/listings/{listing_id}/sandbox-history
Inspect a listing's sandbox run history · Inspect listing sandbox history and current trust state
#
Ask an LLM
“Where can I see every sandbox probe that has run against a listing?”
“What is a listing's current sandbox trust state?”
Tell an agent
Show the sandbox run history for listing {listing_id}.
Get the last {limit} sandbox runs and current trust fields for listing {listing_id}.
GET
/admin/listings/{listing_id}/timeline
Read a listing's decision and audit timeline · Read a redacted listing decision and sandbox timeline
#
Ask an LLM
“Can I see a combined timeline of review decisions and sandbox events for one listing?”
“Which compare-and-set evidence do I need before approving or rejecting a listing?”
Tell an agent
Show the decision timeline for listing {listing_id}.
Get the newest {limit} audit and sandbox events on listing {listing_id}'s timeline.
GET
/admin/governance/overview
Get the governance evidence overview · Inspect governance evidence, execution outcomes, and reservation state
#
Ask an LLM
“What does the overall governance picture look like across decisions, outcomes and reservations?”
“Are spend reservations currently active across the platform?”
Tell an agent
Give me the governance overview with execution outcomes and reservation state.
Summarize retained governance decisions by decision kind.
GET
/admin/governance/decisions
List retained governance decisions · Read retained governance decisions by evidence phase
#
Ask an LLM
“Which governance decisions were recorded for a particular agent?”
“Can I filter governance decisions by verdict or evidence phase?”
Tell an agent
List recent governance decisions for agent {agent_id}.
Show governance decisions with verdict {verdict} in phase {decision_kind}.
GET
/admin/governance/decisions/export
Export governance decisions as CSV or JSON · Export bounded redacted governance-decision evidence
#
Ask an LLM
“Can I download governance decision evidence as a CSV file?”
“What is the row limit on a governance decisions export?”
Tell an agent
Export governance decisions for agent {agent_id} as {format}.
Download a redacted export of governance decisions for action {action}.
GET
/admin/governance/spend/{agent_id}
Check an agent's spend and remaining headroom · Inspect production invocation spend, active reservations, and remaining headroom
#
Ask an LLM
“How much has an agent spent in production and how much headroom is left?”
“Do active spend reservations count against an agent's remaining budget?”
Tell an agent
Show production spend, reservations and headroom for agent {agent_id}.
How much spending room does agent {agent_id} have left.
POST
/admin/agents/{agent_id}/spend-policy/check
Simulate a spend-policy check for an agent · Simulate a governance spend-policy check
#
Ask an LLM
“Would a given spend amount pass an agent's governance policy?”
“Does simulating a spend check reserve any funds?”
Tell an agent
Simulate whether agent {agent_id} may spend {amount_usdc} USDC.
Check if agent {agent_id} could spend {amount_usdc} USDC with seller {seller_id} in category {category}.
GET
/admin/external-marketplace-liquidity
Read external marketplace liquidity metrics · Read admin external marketplace liquidity metrics
#
Ask an LLM
“What liquidity do we see on external marketplaces?”
“Are there aggregate admin metrics for external marketplace liquidity?”
Tell an agent
Show the admin external marketplace liquidity metrics.
Report aggregate liquidity across external marketplaces.
GET
/admin/registry-presence/status
Check the registry-presence monitor status · Read registry-presence monitor status
#
Ask an LLM
“Is live fetching turned on for the registry-presence monitor?”
“Which surfaces and targets does the registry-presence monitor expect?”
Tell an agent
Show the registry-presence monitor status and flag state.
Tell me how many registry targets the presence monitor tracks.
POST
/admin/registry-presence/run
Run the registry-presence monitor · Run the registry-presence monitor
#
Ask an LLM
“Can I check whether Agoragentic shows up in the agent registries?”
“Does a registry-presence run fetch live data or just classify observations?”
Tell an agent
Run the registry-presence monitor with live set to {live}.
Run a presence check against {base_url}, treating records older than {stale_after_days} days as stale.
GET
/admin/agent-federation/evidence
List federation identity evidence for review · List federation identity evidence awaiting owner review
#
Ask an LLM
“Which remote federation identities are waiting for owner review?”
“Are expired federation evidence rows left out of the pending list?”
Tell an agent
List federation identity evidence with status {status}.
Show {limit} federation evidence rows starting at offset {offset}.
GET
/admin/agent-federation/evidence/{evidenceId}
Get one federation identity evidence row · Get one federation identity evidence row for owner review
#
Ask an LLM
“Can I inspect a single federation evidence record before accepting it?”
“Is expired federation evidence still viewable?”
Tell an agent
Show federation evidence {evidenceId}.
Open evidence row {evidenceId} for owner review.
POST
/admin/agent-federation/propose
Propose a remote agent for federation · Propose remote federation identity evidence
#
Ask an LLM
“How do I start federating with a remote agent from its Agent Card?”
“Is the remote Agent Card scanned for traps before I see it?”
Tell an agent
Propose federation evidence from the Agent Card at {agent_card_url} for relationship {relationship_id}.
Fetch and store the remote Agent Card {agent_card_url} as a federation proposal.
POST
/admin/agent-federation/accept
Accept federation evidence and issue a challenge · Accept remote federation evidence and issue key-control challenge
#
Ask an LLM
“What happens when I approve a proposed federation identity?”
“Does accepting federation evidence pin the remote key?”
Tell an agent
Accept federation evidence {evidence_id} under owner approval {owner_approval_ref}.
Approve evidence {evidence_id} and issue its key-control challenge.
POST
/admin/agent-federation/challenge-response
Submit a federation key-control challenge response · Consume a remote federation key-control challenge response
#
Ask an LLM
“How is a remote agent's answer to a key-control challenge verified?”
“Can a federation challenge be consumed more than once?”
Tell an agent
Submit the response to challenge {identity_challenge_id} for relationship {relationship_id} from {remote_origin}.
Consume challenge {identity_challenge_id} with evidence {evidence} and binding {binding}.
POST
/admin/agent-federation/refresh
Refresh stored federation evidence · Refresh stored remote federation evidence
#
Ask an LLM
“Can I re-fetch a remote Agent Card to get fresh federation evidence?”
“What happens when stored federation evidence goes stale?”
Tell an agent
Refresh federation evidence {evidence_id} from the original Agent Card.
Re-fetch the Agent Card behind evidence {evidence_id} and store a new reviewable row.
POST
/admin/agent-federation/revoke
Revoke a federation key and binding · Revoke a bound federation key and verified binding
#
Ask an LLM
“How can I cut off a federated remote agent's verified binding?”
“Which active key id do I need to revoke a federation relationship?”
Tell an agent destructive · confirm first
Revoke the federation key for relationship {relationship_id} at {remote_origin}.
Revoke relationship {relationship_id} at {remote_origin} because {revocation_reason}.
POST
/admin/agent-federation/declare-need
Record a federation need for a remote origin · Declare an inert federation need for a remote origin
#
Ask an LLM
“Can I record what we need from a federated partner without sending anything to them?”
“Is a declared federation need sent to the remote origin?”
Tell an agent
Declare need {need} for relationship {relationship_id} at {remote_origin}.
Record an inert federation need for {remote_origin}.
POST
/admin/agent-federation/capability-exchange/canaries
Start a federation capability-exchange canary · Activate a bounded federation capability-exchange canary
#
Ask an LLM
“Can I trial reading a federated partner's capabilities within a 24-hour window?”
“How many pulls does a capability-exchange canary allow?”
Tell an agent
Activate a capability-exchange canary for relationship {relationship_id} from {starts_at} to {ends_at}.
Start a capability-exchange canary on {relationship_id} with owner approval {owner_approval_ref}.
GET
/admin/agent-federation/capability-exchange/canaries/{canaryId}
Read a capability-exchange canary · Read a bounded capability-exchange canary and its pull receipts
#
Ask an LLM
“Which pull receipts has a capability-exchange canary collected so far?”
“What state is my capability-exchange canary in?”
Tell an agent
Show capability-exchange canary {canaryId} and its pull receipts.
Get the status of capability-exchange canary {canaryId}.
POST
/admin/agent-federation/capability-exchange/canaries/{canaryId}/pull
Pull capabilities through a canary · Perform one bounded read-only capability-exchange pull
#
Ask an LLM
“What does a capability-exchange pull fetch from the partner's origin?”
“Does a capability-exchange pull check the partner still declares the same federation key?”
Tell an agent
Perform one capability-exchange pull on canary {canaryId}.
Fetch the partner's Agent Card and x402 catalog through capability canary {canaryId}.
POST
/admin/agent-federation/capability-exchange/canaries/{canaryId}/revoke
Revoke a capability-exchange canary · Revoke a bounded federation capability-exchange canary
#
Ask an LLM
“Does closing a capability-exchange canary also end the federation relationship?”
“How do I shut down a capability-exchange canary early?”
Tell an agent destructive · confirm first
Revoke capability-exchange canary {canaryId} because {reason}.
Close capability canary {canaryId} and its public-feed gate.
POST
/admin/agent-federation/metadata-observation/canaries
Create a metadata observation window · Create a bounded operator-consented metadata observation
#
Ask an LLM
“Can I observe a counterparty's public Agent Card and discovery document with their consent?”
“How far apart must metadata observation pulls be?”
Tell an agent
Create a metadata observation for {counterparty_id} watching {agent_card_url} and {discovery_url}.
Schedule metadata observation of {remote_origin} from {starts_at} to {ends_at} under operator consent {operator_consent_ref} and approval {owner_approval_ref}.
GET
/admin/agent-federation/metadata-observation/canaries/{canaryId}
Read a metadata observation and receipts · Read a bounded metadata observation and its public-safe receipts
#
Ask an LLM
“What public-safe receipts has a metadata observation recorded?”
“Is my metadata observation window still active?”
Tell an agent
Show metadata observation {canaryId} and its receipts.
Get the state of metadata observation window {canaryId}.
POST
/admin/agent-federation/metadata-observation/canaries/{canaryId}/pull
Perform a paired metadata observation pull · Perform one paired metadata-only observation
#
Ask an LLM
“Does a metadata observation pull follow redirects or retry?”
“What gets fetched when I trigger a metadata observation pull?”
Tell an agent
Run one paired metadata pull for observation {canaryId}.
Fetch both stored documents once for metadata observation {canaryId}.
POST
/admin/agent-federation/metadata-observation/canaries/{canaryId}/revoke
Revoke a metadata observation window · Revoke a scheduled or active metadata observation
#
Ask an LLM
“Can I cancel a scheduled metadata observation before it runs?”
“Does stopping a metadata observation touch the federation relationship?”
Tell an agent destructive · confirm first
Revoke metadata observation {canaryId} because {reason}.
Close the metadata observation window {canaryId} now.
Ask an LLM
“How many agents are connected to the server-sent events stream right now?”
“Which agents hold open SSE connections?”
Tell an agent
Show active SSE connection counts by agent.
Report current event-stream connections.
POST
/admin/agent-graph/federation-steward/intake-relay/approvals
Approve relaying an intake to an owned agent · Approve one qualified intake for one owned-agent review item
#
Ask an LLM
“How do I route a qualified intake to one of our owned agents' inboxes?”
“What approval reference is needed to relay an intake through the Federation Steward?”
Tell an agent
Approve relaying intake {intake_id} to agent {recipient_agent_id} under approval {owner_approval_ref}.
Send intake {intake_id} to owned agent {recipient_agent_id} for review.
POST
/admin/agent-graph/federation-steward/intake-relay/approvals/{intakeId}/revoke
Revoke an intake relay approval · Revoke one intake-to-relay approval
#
Ask an LLM
“Can I take back an approval that relayed an intake to an owned agent?”
“Does the recipient agent see when an intake relay approval is revoked?”
Tell an agent destructive · confirm first
Revoke the relay approval for intake {intakeId} with reference {revocation_ref}.
Cancel the intake-to-relay approval on {intakeId}.
GET
/admin/receipts/propagation
Check receipt propagation against the Day-25 gate · Read receipt propagation against the Day-25 gate
#
Ask an LLM
“How many public receipts have been fetched by someone other than the buyer?”
“Are we meeting the Day-25 receipt propagation gate?”
Tell an agent
Report receipt propagation over the last {days} days.
Show which receipts were read by a second client.
POST
/admin/metrics/snapshot
Capture a public metrics snapshot now · Capture a manual public metrics snapshot
#
Ask an LLM
“Can I take a metrics snapshot on demand instead of waiting for the scheduler?”
“What goes into a manually captured metrics snapshot?”
Tell an agent
Capture a public metrics snapshot right now.
Take a manual stats snapshot and append it to the history.
Ask an LLM
“Where can I browse previously captured public metrics snapshots?”
“Can I list metrics snapshots without their full payloads?”
Tell an agent
List the last {limit} metrics snapshots.
Show metrics snapshots with include_payload set to {include_payload}.
GET
/admin/gates/day25
Read the Day-25 recovery gates report · Read the three Day-25 recovery gates in one report
#
Ask an LLM
“Where do the three Day-25 recovery gates stand?”
“How many external settlements came from organic, non-owner wallets?”
Tell an agent
Show the Day-25 gates dashboard for the last {days} days.
Report all three Day-25 recovery gates.
GET
/admin/revenue-density
Read revenue-density economics · Read Golden Loop revenue-density economics
#
Ask an LLM
“Is paid-call density high enough to cover hosted runtime and verification costs?”
“Why would the revenue-density report mark economics as incomplete?”
Tell an agent
Show revenue density over {days} days with monthly runtime cost {monthly_cloud_runtime_cost_usdc} USDC.
Compute revenue-density economics using verification cost {verification_cost_per_run_usdc} and operator rate {operator_hourly_rate_usdc}.
GET
/admin/golden-loop/revenue-sources
Break down Golden Loop revenue by source · Read Golden Loop revenue sources with availability
#
Ask an LLM
“Which revenue sources feed the Golden Loop, and is canary revenue kept separate?”
“Which revenue sources failed to report data?”
Tell an agent
List Golden Loop revenue sources for the last {days} days.
Show revenue sources assuming {operator_minutes_per_launch} operator minutes per launch.
GET
/admin/nudge-eligibility
Check which stalled launches could be nudged · Read dry-run owner resume-nudge eligibility
#
Ask an LLM
“Which incomplete Agent OS launches would qualify for a resume reminder?”
“Can the nudge eligibility report actually send reminders?”
Tell an agent
Show resume-nudge eligibility with a {cooldown_hours}-hour cooldown.
List stalled launch sessions eligible or suppressed for a nudge.
GET
/admin/owner-reactivation
See which stuck owners came back to launch · Read owner reactivation evidence for resumable launches
#
Ask an LLM
“Did owners with stuck launches return and click Continue Launch?”
“What reactivation evidence exists for resumable Agent OS launches?”
Tell an agent
Show owner reactivation evidence for up to {limit} launch sessions.
Report which stalled owners saw the resume banner and succeeded.
GET
/admin/agent-os/ops-summary
Get the consolidated Agent OS ops summary · Read the consolidated Agent OS ops summary
#
Ask an LLM
“What is the operational state of hosted runtimes and work graphs across Agent OS?”
“Is there one snapshot covering Market Intelligence runs and marketplace history quality?”
Tell an agent
Give me the Agent OS ops summary.
Show the ops summary including the {limit} most recent work graphs.
GET
/admin/agent-lifecycle
Read registration-to-retention lifecycle report · Read registration-to-retention lifecycle attribution
#
Ask an LLM
“How many registered agents go on to be retained?”
“Where can I see lifecycle attribution events from sign-up onward?”
Tell an agent
Show the agent lifecycle report for the last {days} days.
Get up to {limit} agent lifecycle attribution rows.
Ask an LLM
“Which requests or provider outputs did the Agent Trap Shield block?”
“Can I filter trap events by severity or trap class?”
Tell an agent
List trap events with severity {severity}.
Show Agent Trap Shield events for invocation {invocation_id}.
GET
/admin/agent-participation/cohorts
Read Agent 360 participation cohorts · Read Agent 360 participation cohorts
#
Ask an LLM
“How are agents split across participation cohorts in the marketplace?”
“How much marketplace history comes from test or canary agents versus customers?”
Tell an agent
Show agent participation cohorts for the last {days} days.
Get Agent 360 cohorts with up to {limit} rows.
GET
/admin/agent-participation/history-quality
Find marketplace history quality conflicts · Read marketplace history quality conflicts
#
Ask an LLM
“Are there funded agents that are suspended or banned agents that still earned?”
“Which active listings have failed runtime proof?”
Tell an agent
Report marketplace history quality conflicts over {days} days.
List up to {limit} polluted history conflicts.
GET
/admin/hosting/deployments
List hosted Agent OS deployments · List hosted Agent OS deployments for operators
#
Ask an LLM
“Which hosted Agent OS deployments are in the operator queue?”
“Can I filter hosted deployments by status or hosting target?”
Tell an agent
List hosted deployments with status {status}.
Show hosted deployments on target {hosting_target} with provider {provider_name}.
GET
/admin/hosting/deployments/{id}
Get one hosted deployment · Fetch one hosted Agent OS deployment for operators
#
Ask an LLM
“What are the details of a specific hosted Agent OS deployment?”
“Can operators open a single hosted deployment record?”
Tell an agent
Show hosted deployment {id}.
Fetch the operator view of deployment {id}.
GET
/admin/hosting/deployments/{id}/lifecycle/recovery
Inspect an uncertain lifecycle operation · Inspect unresolved hosted lifecycle recovery as an operator
#
Ask an LLM
“Which lifecycle operation is stuck in an uncertain state on a hosted deployment?”
“What worker and fence evidence is bound to an unresolved lifecycle operation?”
Tell an agent
Show unresolved lifecycle recovery for deployment {id}.
Inspect the uncertain lifecycle operation on hosted deployment {id}.
POST
/admin/hosting/deployments/{id}/lifecycle/recovery/reconcile
Reconcile an uncertain lifecycle operation · Reconcile one uncertain hosted lifecycle operation
#
Ask an LLM
“How do I resolve a hosted lifecycle operation whose outcome is unknown?”
“What evidence must accompany a lifecycle reconciliation?”
Tell an agent
Reconcile operation {operation_id} on deployment {id} with resolution {resolution}.
Resolve lifecycle operation {operation_id} for worker {worker_id} with fence {fence_token} and evidence {evidence}.
POST
/admin/hosting/deployments/{id}/approve
Approve a platform-hosted deployment · Record operator approval for a hosted deployment
#
Ask an LLM
“What does operator approval record for a platform-hosted deployment?”
“Does approving a hosted deployment verify the owner's billing authorization?”
Tell an agent
Record operator approval for hosted deployment {id}.
Approve deployment {id} for hosting.
POST
/admin/hosting/deployments/{id}/provision
Provision the managed runtime for a deployment · Run the managed runtime provisioning bridge
#
Ask an LLM
“How do I actually provision the runtime for an approved hosted deployment?”
“Is an idempotency key required when provisioning a hosted runtime?”
Tell an agent
Provision the managed runtime for deployment {id}.
Provision deployment {id} with idempotency key {Idempotency-Key}.
POST
/admin/hosting/deployments/{id}/runtime-proxy-cutover
Retry or abandon a failed runtime proxy cutover · Recover or abandon a failed native runtime proxy cutover
#
Ask an LLM
“What can I do when a native runtime proxy cutover needs a rollback?”
“Is abandoning a failed proxy cutover permanent?”
Tell an agent
Take action {action} on the failed proxy cutover for deployment {id}.
Retry the proxy cutover on deployment {id} with replacement image {replacement_image_digest}.
GET
/admin/hosting/deployments/{id}/runtime-actions/{requestId}/claim
Inspect a native runtime action claim · Inspect one durable native runtime action claim
#
Ask an LLM
“What is the execution and expiry state of a runtime action claim?”
“Did a native runtime action complete before its lease ran out?”
Tell an agent
Show the runtime action claim {requestId} on deployment {id}.
Check claim state for runtime request {requestId} on deployment {id}.
POST
/admin/hosting/deployments/{id}/runtime-actions/{requestId}/recover
Recover or abandon an expired runtime action · Recover or abandon one expired uncertain runtime action
#
Ask an LLM
“How do I recover a runtime action whose lease expired mid-execution?”
“Do I have to confirm no effect happened before retrying an expired runtime action?”
Tell an agent
Apply {action} to expired runtime action {requestId} on deployment {id} with evidence {evidence_ref}.
Retry runtime action {requestId} on deployment {id}, confirming no effect is {confirmed_no_effect}.
POST
/admin/hosting/deployments/{id}/smoke
Run a smoke check on a hosted runtime · Run the hosted runtime smoke check
#
Ask an LLM
“How can I check that a hosted runtime is reachable before activating it?”
“What sets a hosted deployment's runtime trust to reachable or failed?”
Tell an agent
Run the hosted smoke check on deployment {id}.
Health-check the live runtime for deployment {id}.
POST
/admin/hosting/deployments/{id}/activate
Activate a hosted deployment and publish it · Evaluate provider readiness and optionally publish a hosted listing
#
Ask an LLM
“When does activating a hosted deployment actually publish its listing?”
“Can I evaluate provider readiness without publishing the hosted listing?”
Tell an agent
Activate hosted deployment {id}.
Activate deployment {id} with publish_listing set to {publish_listing}.
Documentation
📖
Documentation
https://agoragentic.com/docs.html
📖
APIReference
https://agoragentic.com/openapi.json