Dynamic Feed Agent API
The agent API from Dynamic Feed — 53 operation(s) for agent.
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
Ask an LLM
“Which Dynamic Feed tool should my agent use for live rainfall data in a given city?”
“Can I search capabilities in plain words and see measured uptime and latency for each match?”
Tell an agent
Discover capabilities that match the need {need}.
Find the top {limit} capabilities for {need} and leave out external MCP servers.
Ask an LLM
“Is there a POST version of capability discovery that takes the need as a JSON body?”
“Can my agent send its capability-discovery request as JSON instead of query parameters?”
Tell an agent
Send a JSON body to the POST discover endpoint asking for sanctions screening capabilities.
Post a discovery request in JSON for satellite position tools including external servers.
Ask an LLM
“What receipt schemas does Dynamic Feed publish, and which mint endpoint goes with each?”
“Are deprecated or quarantined receipt schemas still listed in the schema index?”
Tell an agent
Show me the DF-VERIFY/1 receipt schema index.
List every receipt schema profile with its $id and title.
Ask an LLM
“What curated tool bundles exist so I can pick tools for my agent's use case?”
“Are there ready-made groupings of data tools per domain like finance or weather?”
Tell an agent
List all curated tool bundles.
Show me every domain bundle available for agents.
Ask an LLM
“Which tools are in the devsecops bundle?”
“Can I look up a single tool bundle by its id to see what it includes?”
Tell an agent
Show me the tools in bundle {bundle_id}.
Open the {bundle_id} bundle and list its tools.
Ask an LLM
“Can I fetch weather and earthquakes together in a single free request?”
“How many tool calls can I bundle into one keyless batch request?”
Tell an agent
Run a free batch that gets current weather for London and recent earthquakes.
Get the list of tool names available for the keyless batch endpoint.
Ask an LLM
“Can my agent pay per call in USDC for one signed datapoint without signing up?”
“Is there a premium single-call endpoint that returns an Ed25519-signed result?”
Tell an agent
Pay for one signed call to the {tool} tool via x402.
Buy a single premium signed datapoint from {tool}.
Ask an LLM
“What does a single paid call to a given tool cost, and what asset is it paid in?”
“Can I probe a pay-per-call tool without being charged to learn how to pay?”
Tell an agent
Show me the x402 payment challenge for the {tool} tool without paying.
Check the price of a paid single call to {tool}.
Ask an LLM
“Can I pay per call to replay what a tool reported at a past time without an API key?”
“Is there a keyless paid way to get a signed archived observation from a specific moment?”
Tell an agent
Pay via x402 to replay archived current weather as of a specific past timestamp.
Buy a signed as-of replay from the archive without using an API key.
Ask an LLM
“Can my agent pay with a wallet to run more than 20 tool calls in one request?”
“What is the cap on signed tool calls in the paid batch tier?”
Tell an agent
Pay via x402 for a premium batch of up to 100 signed tool calls.
Run a paid signed batch of data tool calls from my agent's wallet.
Ask an LLM
“What paid endpoints does Dynamic Feed sell and how much does each cost?”
“Which facilitator settles x402 payments, and which wallet receives them?”
Tell an agent
Show me the x402 price list for every paid endpoint.
List the pay-per-call SKUs with their input and output schemas.
Ask an LLM
“Can I check whether facts an LLM stated, like the latest Python version, are out of date?”
“Does drift detection grade each stale claim by severity and give a correction?”
Tell an agent
Audit the claims 'the latest Python is 3.11' and 'Node.js is on 18' for drift.
Check this LLM answer for facts that have drifted from current ground truth.
Ask an LLM
“Can I screen my chatbot's draft answer for hallucinated facts before the user sees it?”
“What does the pre-answer guard return when a draft needs fixing?”
Tell an agent
Run the pre-answer guard on my AI's draft reply and tell me pass or revise.
Get the corrections needed for this draft answer before I send it.
Ask an LLM
“Which upstream sources back Dynamic Feed's tools, and can I use them commercially?”
“What licence and update cadence does each data feed have?”
Tell an agent
List every upstream source with its publisher, licence and update cadence.
Show me which sources permit commercial use and redistribution.
Ask an LLM
“Where does the earthquake data come from, and what is its chain of custody?”
“Can I trace a number back to its upstream source and licence with a simple GET?”
Tell an agent
Explain the source and custody trail for {subject} using the GET endpoint.
Trace where the {subject} data originates.
Ask an LLM
“Can I post a JSON subject to get its source provenance instead of using a query string?”
“Is there a POST form of the chain-of-custody explanation?”
Tell an agent
Post the subject {subject} to the explain endpoint for its provenance.
Use the POST explain call to show the licence and custody chain for {subject}.
Ask an LLM
“What did Dynamic Feed report about the world on a given date?”
“What happens if I request a snapshot from before the archive began?”
Tell an agent
Get the world snapshot stored for {date}.
Capture today's world-state snapshot.
Ask an LLM
“What did the weather tool say at the exact moment I made a decision last week?”
“Does the as-of replay re-fetch data or return exactly what was served then?”
Tell an agent
Replay what {tool} reported as of {t}.
Show the archived {tool} value nearest before {t}.
Ask an LLM
“Which subjects am I monitoring on my watchlist?”
“Can I pull current values for everything on my watchlist in one call?”
Tell an agent
Show my saved watchlist.
Show my watchlist with pull set to {pull} so each item's live value is included.
Ask an LLM
“Can I add BTC to my watchlist so I can monitor it?”
“What is the maximum number of items a watchlist can hold?”
Tell an agent
Add BTC via the cite tool to my watchlist.
Replace my entire watchlist with a new set of items.
Ask an LLM
“Can I remove a single item from my watchlist by its label?”
“What happens if I delete from my watchlist without giving a label?”
Tell an agent destructive · confirm first
Remove {label} from my watchlist.
Clear my whole watchlist.
Ask an LLM
“Can I subscribe to a live value with Server-Sent Events instead of polling?”
“What interval and update count limits apply to the live stream?”
Tell an agent
Stream {tool} with args {args} every {interval} seconds.
Open a live stream of {tool} and stop after {count} updates.
Ask an LLM
“Can I get a webhook call whenever the exploited vulnerabilities list changes?”
“How are webhook deliveries signed so I can verify they are genuine?”
Tell an agent
Register a webhook to my endpoint that fires when a tool's value changes.
Create a new webhook subscription for exploited vulnerability changes.
Ask an LLM
“Which webhook subscriptions have I registered?”
“Are webhook secrets returned when I list my subscriptions?”
Tell an agent
List my webhook subscriptions.
Show all webhooks I have set up.
Ask an LLM
“Can I stop a webhook subscription I no longer need?”
“What do I need to pass to delete one of my webhooks?”
Tell an agent destructive · confirm first
Delete webhook subscription {id}.
Remove my webhook with id {id}.
Ask an LLM
“Can I send a test delivery to check my endpoint validates the HMAC signature?”
“Is there a way to trigger a webhook now without waiting for a value to change?”
Tell an agent
Send a signed test delivery to my webhook subscription.
Fire a test event at one of my webhooks now.
Ask an LLM
“What capabilities does the v1 agent and robot API offer at a glance?”
“Is there a single page listing every v1 endpoint with copy-paste examples?”
Tell an agent
Show me the v1 API index.
List every v1 capability with an example call.
Ask an LLM
“Can I get timestamped, signed proof of what my AI was told?”
“Which key do I verify an evidence receipt's signature against?”
Tell an agent
Wrap this claim in a signed evidence receipt.
Create an Ed25519-signed receipt for this JSON data.
Ask an LLM
“Can I create an independently verifiable record of some JSON without an API key?”
“Does a stamp signature prove the JSON content is true or only its integrity?”
Tell an agent
Stamp this JSON document as a verifiable record.
Create an integrity stamp for my submitted JSON.
Ask an LLM
“Can I prove after the fact what inputs my AI agent acted on and what it decided?”
“What fields go into an agent receipt besides the inputs?”
Tell an agent
Create an agent receipt for my agent's last action and its inputs.
Record a signed receipt of the decision my agent made.
Ask an LLM
“Can a neutral third party notarize my data feed so I can prove it existed at a moment?”
“Can I notarize only a SHA-256 hash and keep the data private?”
Tell an agent
Notarize this JSON payload.
Notarize my data by sending only its SHA-256 hash.
Ask an LLM
“Where do my organisation's decisions rely on external data we cannot verify?”
“Can I get a signed exposure report for specific perils at a location?”
Tell an agent
Build an exposure report for {organisation} covering {perils} at {place}.
Generate an exposure report for the {use_case} use case at {lat}, {lon}.
Ask an LLM
“Can I wrap a wildfire detection with independent fire-weather and smoke context?”
“What context is captured in a wildfire chain-of-custody evidence record?”
Tell an agent
Create a wildfire evidence record for a detection at {lat}, {lon}.
Build wildfire evidence at {lat}, {lon} detected by {detector} at {detected_at}.
Ask an LLM
“Is the legacy verifiable-credential wrapping endpoint still safe to use?”
“Why was the VC wrapping endpoint put in security quarantine?”
Tell an agent
Call the quarantined legacy VC wrapping endpoint.
Submit a receipt to the legacy verifiable credential wrapper.
Ask an LLM
“Can I register my own weather sensor so its readings are witnessed as signed receipts?”
“Does registering a station mean Dynamic Feed measures the values itself?”
Tell an agent
Register my sensor station with its name, location and quantities.
Sign up a new station that reports wind speed.
Ask an LLM
“Can I get a signed receipt for a value my registered station reported?”
“Is a station reading receipt marked as self-reported rather than measured?”
Tell an agent
Witness a wind reading from my registered station.
Submit my station's latest reading for a signed receipt.
Ask an LLM
“What does a registered station report, without revealing who owns it?”
“Can a verifier look up the station behind a station receipt?”
Tell an agent
Show the public descriptor for station {station_id}.
Look up what station {station_id} reports.
Ask an LLM
“Can I route a third-party API call through a witness and get a signed record of its response?”
“Are private network addresses blocked when proxying a call for witnessing?”
Tell an agent
Fetch this third-party URL through the proxy witness and sign what it returned.
Get a signed archived receipt of what an external API responded right now.
Ask an LLM
“Can a neutral party sign and timestamp a machine-to-machine transaction record?”
“Does the transaction witness ever hold or move funds?”
Tell an agent
Witness this agent-to-agent transaction record.
Get a signed, logged receipt that this transaction record exists and is intact.
Ask an LLM
“What does a transaction witness receipt look like before I integrate?”
“Is there a demo txn-receipt I can verify in the browser?”
Tell an agent
Show me a sample transaction witness receipt.
Get the demo txn-receipt/v1 to test verification.
Ask an LLM
“Can I prove which model served an inference without sending the prompt or response?”
“What hashes and metadata go into an inference receipt?”
Tell an agent
Create an inference receipt from my prompt and response hashes.
Witness this AI inference using only its SHA-256 hashes.
Ask an LLM
“What does an inference receipt look like when a provider served a different model?”
“Is there a sample inference receipt I can verify in my browser?”
Tell an agent
Show me a sample inference receipt.
Get the demo inference-receipt/v1.
Ask an LLM
“Can I get a signed Verified Eval stamp while keeping my test set private?”
“How does committing a salted hash of my test set prevent retuning after the run?”
Tell an agent
Create an eval receipt for my benchmark run with its set commitment.
Witness this AI evaluation run as a Verified Eval.
Ask an LLM
“What does a Verified Eval receipt look like?”
“Is there a demo eval receipt I can check before submitting my own run?”
Tell an agent
Show me a sample eval receipt.
Get the demo eval-receipt/v1 for verification testing.
Ask an LLM
“Can I verify the archive is append-only by checking its daily digests?”
“What is the anchor status of recent archive checkpoints?”
Tell an agent
List recent daily archive checkpoints.
Show the latest {limit} archive checkpoints.
Ask an LLM
“Can I download the RFC 3161 token for a given day's archive checkpoint?”
“Which proof formats are available for an archive checkpoint?”
Tell an agent
Download the checkpoint proof for {day}.
Get the {kind} proof for the {day} archive checkpoint.
Ask an LLM
“What is the current head hash and entry count of the notary log?”
“Which signed artifact should I anchor to detect tampering with past notarizations?”
Tell an agent
Show the current notary log head.
Get the signed chain head hash of the transparency log.
Ask an LLM
“Can I fetch a single notarization log entry by its index?”
“How do I recompute an entry hash to verify chain linkage?”
Tell an agent
Show notary log entry {idx}.
Fetch transparency log entry number {idx}.
Ask an LLM
“Can I browse all notarizations in the public transparency log?”
“Does the transparency log contain payloads or only their hashes?”
Tell an agent
List the newest notarizations in the transparency log.
Page the notary log back from entry {before}, {limit} at a time.
Ask an LLM
“Can I prove a notarization is part of the anchored history without trusting Dynamic Feed?”
“What does a notary log inclusion proof chain back to?”
Tell an agent
Get the inclusion proof for notary entry {idx}.
Prove that log entry {idx} is in the anchored history.
Ask an LLM
“Is there a compact Merkle audit path I can verify in my browser for a witness entry?”
“Can I check a witness entry's inclusion with only the published key?”
Tell an agent
Get the Merkle inclusion path for witness entry {idx}.
Show the signed-checkpoint audit path for witness {idx}.
Ask an LLM
“Can I have my feed URL fetched on a schedule and every change notarized automatically?”
“What is the minimum interval and how many active feeds can one key register?”
Tell an agent
Register my feed URL for continuous notarization every 180 minutes.
Start notarizing every change to my company's published feed.
Ask an LLM
“Which feeds have I registered for continuous notarization?”
“When will my notarized feeds next be fetched?”
Tell an agent
List my registered notarization feeds.
Show the status and last hash of each feed I notarize.
Ask an LLM
“Is there a benchmark of facts that change after a model's training cutoff?”
“Can I grade a model without tools against answers measured from live sources?”
Tell an agent
Get the current live-facts benchmark items.
Show me the signed live-facts eval set with grading rules.
Ask an LLM
“How fast do live-fact benchmark answers drift day over day?”
“Can I diff past daily snapshots of the live-facts benchmark?”
Tell an agent
Show live-facts benchmark snapshots for the last {days} days.
Get the full daily history of the live-facts benchmark.
Ask an LLM
“What earthquakes, wildfires or disasters are happening near my drone's location?”
“Can I limit a hazard scan to a radius in kilometres?”
Tell an agent
Scan for hazards near {lat}, {lon}.
List live hazards within {radius_km} km of {lat}, {lon}.
Ask an LLM
“Can I get a signed MET or NOT-MET record for wind speed above a threshold at a place?”
“Which metrics and comparison operators does the parametric attestation support?”
Tell an agent
Attest whether {metric} is {op} {threshold} in {city}.
Check if {metric} exceeds {threshold} at {lat}, {lon} and sign the result.
Ask an LLM
“Can I get a dispute-proof receipt that a parametric insurance trigger fired?”
“Does a trigger receipt pin the methodology so it cannot change after the event?”
Tell an agent
Issue a trigger receipt for {metric} {op} {threshold} at {city} under contract {contract_ref}.
Create a {peril} trigger receipt for {metric} above {threshold} at {lat}, {lon}.
Ask an LLM
“What is the river gauge reading at an exact coordinate for a flood trigger?”
“Which official gauge is used for the intensity at my location?”
Tell an agent
Get the flood intensity at {lat}, {lon}.
Read {metric} for peril {peril} at {lat}, {lon}.
Ask an LLM
“Are there satellite bushfire hotspots near me in Australia?”
“Can I see recent fire hotspots for a single Australian state?”
Tell an agent
Show bushfire hotspots in {state}.
List bushfire hotspots within {radius_km} km of {lat}, {lon}.
Ask an LLM
“How many calls have I used today and how many remain on my key?”
“Which plan is my API key on?”
Tell an agent
Show my API key usage and remaining quota.
Check my plan and today's call count.
Documentation
📖
Documentation
https://dynamicfeed.ai/docs
📖
APIReference
https://dynamicfeed.ai/docs
📖
GettingStarted
https://dynamicfeed.ai/connect
📖
RateLimits
https://raw.githubusercontent.com/api-evangelist/dynamicfeed-ai/refs/heads/main/rate-limits/dynamicfeed-ai-rate-limits.yml
Specifications
Schemas & Data
📊
JSONSchema
https://raw.githubusercontent.com/api-evangelist/dynamicfeed-ai/refs/heads/main/json-schema/dynamicfeed-ai-schemas-index.json
📊
JSONSchema
https://dynamicfeed.ai/schemas/index.json
Other Resources
🔗
DataModel
https://raw.githubusercontent.com/api-evangelist/dynamicfeed-ai/refs/heads/main/data-model/dynamicfeed-ai-data-model.yml
🔗
ErrorCatalog
https://raw.githubusercontent.com/api-evangelist/dynamicfeed-ai/refs/heads/main/errors/dynamicfeed-ai-problem-types.yml
🔗
Conventions
https://raw.githubusercontent.com/api-evangelist/dynamicfeed-ai/refs/heads/main/conventions/dynamicfeed-ai-conventions.yml
🔗
Playground
https://dynamicfeed.ai/playground