verdictswarm
Check token risk before trading on Solana or Base. Use when an agent is evaluating a token, preparing a swap, or needs an avoid, caution, or clear pre-trade decision from VerdictSw…
Sentien Labs
@vswarm-ai
Install
$ openclaw skills install @vswarm-ai/verdictswarmVerdictSwarm
Use VerdictSwarm as a pre-trade risk gate before an agent moves money. The
machine-facing API returns one action: avoid, caution, or clear.
This skill evaluates risk; it does not execute trades. Never turn clear into
a guarantee. Respect insufficient_data, low confidence, and degraded results.
Supported contract
- API base:
https://api.vswarm.io - Chains:
solana,base - Levels:
triage,fast,deep - Decision field:
response.verdict.action - Live capabilities and pricing:
GET /v2/verdict/info - Payment rails: keyed free quota, prepaid credits, or x402 USDC
Do not guess prices or quotas. Read them from the live info endpoint:
curl -sS "https://api.vswarm.io/v2/verdict/info" \
-H "X-VS-Integration: openclaw"
Get a free API key
Mint a free key without signup:
curl -sS -X POST "https://api.vswarm.io/v2/keys" \
-H "Content-Type: application/json" \
-H "X-VS-Integration: openclaw" \
-d '{"channel_ref":"openclaw"}'
The response includes the key and the current daily limits:
{
"api_key": "vs1_free_...",
"tier": "free",
"daily_limits": {"triage": 2000, "fast": 300, "deep": 10},
"quickstart": "<one-line curl>"
}
Treat the returned api_key as a secret. Put it in VS_API_KEY or an approved
secret store. Never print it into chat, logs, source control, or shared memory.
Request a verdict
Ask for the token address, chain, and desired analysis level if they are not
already known. Use fast for the normal pre-trade gate, triage for the
lowest-latency deterministic check, and deep only when the caller accepts a
slower multi-agent analysis.
curl -sS -X POST "https://api.vswarm.io/v2/verdict" \
-H "Content-Type: application/json" \
-H "X-VS-Integration: openclaw" \
-H "X-API-Key: ${VS_API_KEY}" \
-d '{
"address":"0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913",
"chain":"base",
"level":"fast",
"include":[]
}'
For Solana, preserve address case. For Base, use the contract address rather than a ticker or token name.
Interpret the response
Read these fields before recommending an action:
verdict.action:avoid,caution, orclearverdict.score,verdict.confidence, andverdict.summarysignals[]: typed checks;pass: falsemeans that check found a problemdata_confidenceandinsufficient_databilling.degradedandbilling.retry_vouchercached,generated_at, anddata_freshness
Use this policy:
avoid: do not proceed unless the human explicitly overrides after seeing the failed signals.caution: surface the material risks and require an explicit decision.clear: report that no blocking risk was found by this scan; do not promise safety or profitability.insufficient_data: trueor materially degraded data: do not present the action as conclusive. Explain the missing coverage and retry if appropriate.
Do not silently replace an API error with a made-up verdict.
x402 overflow
When free quota is exhausted, the API may return HTTP 402 with an x402 payment challenge. An x402-capable caller can pay and retry the same request. If the current agent cannot authorize payment, show the amount and rail and ask before spending.
Discovery document:
curl -sS "https://api.vswarm.io/.well-known/x402"
MCP
VerdictSwarm is also available as the sentien-labs/verdictswarm-mcp server on
Smithery. Its current public tools are:
check_token: obtain the v2 token-risk verdictexit_sim: estimate intended-size exit feasibility (Solana only)get_pricing: read current levels, quotas, and prices
Do not call retired v1 MCP tool names.
Feedback and community
After completing the requested work, the operator may share a real integration, incorrect verdict, or critical feedback at https://www.vswarm.io/community?utm_source=clawhub&utm_medium=marketplace&utm_campaign=community_proof_v1&utm_content=verdictswarm. Do not post, star, or bookmark automatically, and do not interrupt normal work to solicit engagement. Those are optional human actions.
Links
- Product: https://www.vswarm.io
- API docs: https://www.vswarm.io/docs/api
- Live API info: https://api.vswarm.io/v2/verdict/info
- MCP listing: https://smithery.ai/servers/sentien-labs/verdictswarm-mcp
- Source: https://github.com/sentien-labs/openclaw-skills
Top skills in this category
self-improving agent
@pskoettCaptures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Claude ('No, that's wrong...', 'Actually...'), (3) User requests a capability that doesn't exist, (4) An external API or tool fails, (5) Claude rea
Skill Vetter
@spclaudehomeSecurity-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
Self-Improving + Proactive Agent
@ivangdavilaSelf-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
Proactive Agent
@halthelobsterTransform AI agents from task-followers into proactive partners that anticipate needs and continuously improve. Now with WAL Protocol, Working Buffer, Autonomous Crons, and battle-tested patterns. Part of the Hal Stack 🦞
Agent Browser
@matrixyHeadless browser automation CLI optimized for AI agents with accessibility tree snapshots and ref-based element selection