jepeta.
agent api

What Base token risk API can an AI agent use before execution?

An AI agent should call a deterministic, machine-readable risk endpoint with an exact Base contract address and treat the result as policy input rather than natural-language advice. Jepeta exposes a canonical V5 POST public preview, supported V4 GET compatibility, OpenAPI, agent manifests, methodology, a public signal feed, and permanent evidence URLs.

By Jepeta EditorialPublished Updated Reviewed 3 min read
Concept illustration of automated data connections passing through an inspection gate.
Jepeta · Original AI concept illustration; not a product photo or evidence

Key takeaways

  • Validate chain 8453 + address
  • Call canonical V5 POST preview
  • Fail closed on errors

What an agent needs

The integration needs a stable chain identifier, exact-contract validation, bounded schemas, explicit error behavior, and fields that distinguish risk from evidence quality. The agent should not infer a PASS from a network failure or missing upstream data.

Jepeta's canonical V5 preview returns a PASS/WARN/BLOCK verdict with normalized evidence, logical signals, source status, evidence quality, conflicts, and freshness. V4 GET remains available for compatibility.

Execution pattern

Validate the address locally, call the preview endpoint, and reject non-200 or schema-invalid responses. For V5, require status=VERDICT and a non-null PASS, WARN or BLOCK value in the HTTP 200 response. If an error response contains NO_VERDICT, stop and inspect the failure details. Stop on BLOCK, apply additional review policy on WARN, and enforce independent execution limits even on PASS.

For traceability, an agent can use the public evidence catalog when a permanent published snapshot exists. The live preview remains the correct input for a current trade.

Discovery surface

Jepeta publishes OpenAPI, agent.json, llms.txt, llms-full.txt, methodology.json, changelog.json, a recent risk feed, and stable evidence pages. These give both agent runtimes and search/AI crawlers multiple consistent ways to resolve the same product identity.

The scanner and full risk report are free. A permanent evidence page preserves a selected historical snapshot; it is not the same request or freshness contract as a newly generated report.

Give the agent observations, not signing authority

An integration can expose the risk lookup as a read-only tool while leaving the wallet outside that tool. Return the requested chain and contract with the response, validate them before interpreting the verdict, and keep execution approval in a separate policy. A prompt asking the agent to ignore NO_VERDICT must not change the parser or authorize a signature. This is an integration design derived from the API contract, not a claim that Jepeta operates or tested your agent.

AI-assisted review of linked documentation; no independent product benchmark. Screening observations and policy labels do not guarantee token safety.

Practical checklist

  • Validate chain 8453 + address
  • Call canonical V5 POST preview
  • Fail closed on errors
  • Stop on BLOCK
  • Review WARN
  • Apply own policy on PASS
  • Use evidence URLs for traceability

Limitations

This is educational risk-screening content, not investment advice. Token and market conditions can change after any snapshot. A PASS result or negative honeypot signal is not a safety guarantee.

Sources & methodology

Jepeta OpenAPIJepeta agent manifestJepeta public evidence catalogJepeta methodology

Sources reviewed Nov 4, 2026.

Try the public API

An agent can make this single read-only request for the Base AERO contract before evaluating an action. The request connects no wallet and submits no trade. Reject HTTP failures, inspect evidence quality and timestamps, and stop on NO_VERDICT. PASS does not authorize a trade.

curl --fail-with-body --max-time 30 \
  https://aitgmgfumsdqecmanrab.supabase.co/functions/v1/jepeta-risk-scan \
  -H "Content-Type: application/json" \
  --data '{"tokenAddress":"0x940181a94a35a4569e4529a3cdfb74e38fd98631"}'

Request and response fields: canonical V5 OpenAPI specification. See the API documentation for integration limits. No API key is required for this public example.

Continue reading

Browse all articles
Concept illustration of automated data connections passing through an inspection gate.
agent api

How do I add a pre-trade risk gate to a Base trading bot?

Put the risk gate before transaction construction or signing: resolve the exact Base token, call a deterministic risk API, validate the schema, and convert the result into explicit stop/review/continue rules. Do not let an LLM free-form its way around missing evidence.

Nov 4, 2026Read article
Concept illustration of visible evidence in front of obscured, unknown information.
discovery

Is this Base token safe?

No scanner can prove that a token is safe. A useful pre-trade check should instead ask whether the exact Base contract shows known risk signals now: honeypot behavior, mintability, unusual taxes, weak liquidity, concentrated holders, incomplete source data, or other reasons to stop and review. When the required evidence is usable, Jepeta can issue PASS, WARN, or BLOCK. If required evidence is unusable, it returns NO_VERDICT. PASS is never a safety guarantee.

Nov 4, 2026Read article
Concept illustration of a permission gate between two tokens before a swap.
workflow

What should I check before buying a Base token?

Before a Base token trade, verify the exact contract, then check honeypot status, mintability, taxes, holder concentration, liquidity, observed LP-lock evidence, data quality, and source freshness. A single green metric is not enough; the goal is to catch reasons to stop or review before execution.

Nov 4, 2026Read article