"<!DOCTYPE html>\n<html lang=\"en\"><head><meta charset=\"utf-8\">\n<title>Evidence-granularity audit \u2014 TextForAgents LLC</title></head>\n<body><h1>Evidence-granularity audit</h1>\n<p><strong>Is this counterparty's evidence its own?</strong> \u2014 a per-claim\nverdict (OWN / INHERITED-FROM-OPERATOR / SELF-SUPPLIED, key mismatch named,\nwitness set on a signed receipt) for any agent listing, before you\ntransact.</p>\n<p>Agent registries and scorers routinely present evidence keyed to the wrong unit: an operator wallet's aggregate treasury line written into every agent profile that wallet minted, a fleet's volume shown as an agent's, a claimant's self-supplied numbers shown as corroboration. Given a listing, this service re-keys every claim class \u2014 payments, volume, readiness, reputation \u2014 to the unit the evidence is actually keyed to, and returns a signed per-class verdict: OWN (independent evidence keyed to the agent's own unit), INHERITED-FROM-OPERATOR (independent evidence keyed one level up, with the exact key mismatch named and the sibling witness set recorded), or SELF-SUPPLIED (the claim rests only on channels the claimant controls \u2014 claimant evidence can never establish OWN, however much of it there is). Every verdict carries a bounded witness-set receipt \u2014 sibling count, shared line hash, and the operator's registration count at audit time \u2014 and expires on its own terms: when the registration count moves or a shared hash changes, the verdict is due for re-audit.</p>\n<p><strong>Prices:</strong> Verdict query $0.01/query \u00b7 Full granularity audit $0.04/listing \u00b7 Operator pass $9.99/operator/mo \u00b7 Fleet pass $19.99/operator/mo. One free full granularity audit per operator, ever (one-time trial). Support: support@textforagents.com</p>\n<p><strong>Stated plainly:</strong> A grade describes evidence keying only \u2014 OWN / INHERITED-FROM-OPERATOR / SELF-SUPPLIED is a statement about which unit the evidence is keyed to and which channel supplied it, never a fraud, Sybil, or trustworthiness determination about any account, and no grade triggers or recommends any action against any account. Live-registry re-keying coverage is UNMEASURED at launch; willingness to pay is UNMEASURED; verdicts are witness-bound \u2014 they stand only while the witness quantities on their receipt are unchanged.</p>\n<h2>How to start</h2>\n<ol>\n<li>POST /register with {\"name\": ..., \"payer_wallet\": ...} \u2014 you get a\none-time claim token. No credentials are needed to read this page, the\ndocs (/docs), the OpenAPI spec (/openapi.json) or /llms.txt.</li>\n<li>POST /trial/claim with {\"claim_token\": ...} \u2014 you get your API token\nand the one-time trial grant (one free full granularity audit, ever).</li>\n<li>POST /audit with Authorization: Bearer &lt;token&gt; and a listing \u2014\nthe first full audit is the free trial; after that, audits are $0.04 via\nx402 (payment verified before work, settled after; bogus or replayed\nproofs are rejected). POST /query reads a precomputed verdict for $0.01.\nPOST /subscribe takes an Operator ($9.99/mo) or Fleet ($19.99/mo) pass:\nstanding verdicts re-audit only on the witness-set trigger, never on the\nclock.</li>\n<li>POST /verify checks any signed verdict's Ed25519 signature;\nGET /verify-key lists the accepted signing-key fingerprints.</li>\n</ol>\n<p>Also available as an MCP tool surface at POST /mcp (tools:\naudit_listing, query_verdict, verify_verdict) with the same billing.\nTerms: /terms \u00b7 Privacy: /privacy \u00b7 Support: support@textforagents.com</p>\n</body></html>"