Family A — Agent Incident & Trace Assurance
Your agent did something in production nobody can fully explain.
The logs are there. The story is not. This turns a raw agent trace into an independent, exported diagnostic — what happened, in what order, with what evidence — that you can hand to someone outside your team.
Opens your email app with the intake questions pre-filled. There is no form processor on this page.
Fit / not-fit
This fits if
- You run agents in production, not a prototype.
- A real incident occurred, or one recurs.
- You need a finding for someone outside your own team.
- You have trace or log data, even if incomplete.
This is not a fit if
- You want a monitoring or observability platform.
- You want the agent code fixed as the primary service.
- You have no trace data to examine.
Offer and current pricing
| Offer | What it is | Price | Status |
|---|---|---|---|
| Exported agent-trace diagnostic | Primary door for this family — the engagement described above. | $7,500 list $5,250 founding (first three paid Band A engagements; none recorded yet) $12,500 Band B (multi-agent or counsel / claims-facing write-up) |
Offered — scoped by reply |
| Related expansion work | Coding-agent post-mortem, release/change assurance, or continuous trace assurance — only after this door fits. | Scoped after fit | Scoped on request |
Prices are the accepted scope grid. Nothing here was invented for this page. Band B is multi-agent or counsel / claims-facing work, not a discount tier.
What the finding looks like
Illustrative only — no real client, no real incident, no protected methodology:
Trace review covered N tool-call steps across the incident window. What the trace supports: the agent's final action and its immediately preceding tool call. What it does not support: intent behind a specific parameter choice two steps earlier — the trace did not capture that reasoning. What could not be determined: whether an upstream service returned degraded data, because that service's own logs were not in scope for this engagement.
Limitations
After a fit reply: Statement of Work
Once a short exchange confirms fit, you receive a specific Statement of Work — not a generic template dressed up:
- Scope: the exact incident and artifact set to be reviewed.
- Fee: list $7,500; founding $5,250 for the first three paid Band A engagements (none recorded yet); Band B $12,500. The SOW states the applicable figure.
- No success fee. No "no finding, no fee."
- Work starts after you accept the SOW in writing.
The SOW template exists. Sending it to a specific prospect is a human step for that engagement.
Payment — after an accepted SOW
This page is not a storefront checkout. Intended path: fit conversation → accepted SOW → payment instruction for that engagement.
Payment instructions are supplied after an accepted SOW. A Contra one-time payment link for founding Band A ($5,250) exists and is sent per-engagement after SOW acceptance; it is not published on this site. Identity is verified and receiving rails are present. Guest checkout without a Contra account has been verified, with card, US bank (ACH), and crypto visible. A real charge settling, refund or dispute handling, and payout / withdrawal after funds arrive are not yet exercised. A signed SOW does not by itself prove those last steps.
List $7,500 and Band B $12,500 still use a direct invoice (intended US PayPal Invoice in USD; Payoneer backup). That invoice rail is also not proved end-to-end.