- Home
- Skills
- Workflow Automation
- Verify Before You Report — Server-State Truth Checking
Works with the AI Agents you already use
Verify Before You Report — Server-State Truth Checking
Stop your agent from reporting 'submitted' and 'done' for work that never landed.
$29
Verify Before You Report — Server-State Truth Checking
Example session with this skill installed
Verify the status of the 5 user profile updates I just attempted in the staging DB and give me the final report.
- Read your context and instructions
- Compiled the verify before you
I have completed the verification of the 5 profile updates against the authoritative staging database records. I performed a read-back query for each specific user ID to confirm the status and updated_at fields match the intended values.
Verification Report: User Profile Batch
| Status | Count | Details |
|---|---|---|
| Verified | 3 | IDs: USR-001, USR-002, USR-004. Values match intended state. |
| Not yet landed | 1 | ID: USR-005. Record exists but status still shows PENDING. |
| Contradicted | 1 | ID: USR-003. Server record shows update_error: validation_failed. |
The claim for USR-003 is withdrawn as the server-side record contradicts the local success log. USR-005 is currently in an asynchronous processing state and cannot be marked as complete.
Next steps
- Inspect the validation logs for
USR-003to resolve the field contradiction. - Re-probe
USR-005in 60 seconds to check if the background process has settled. - Proceed with downstream notifications only for the 3 verified records.
Connects securely to your tools. The creator never sees your data.
About this skill
Every agent operator has lived this: the script log said success, the sub-agent said done, the API returned 200 — and the record was never actually saved. Then the follow-up actions, the client update, and the duplicate submission all got built on air.
This skill installs a verification discipline in your agent:
- Treats every local signal — editor views, run logs, worker reports, success responses — as an unverified claim, never as proof
- Names the authoritative server-side record for each system before work starts
- Probes per record by stable ID, re-probes after interruptions (work often lands exactly when a run stops — blind retries create duplicates)
- Reads back every critical write and matches it field by field
- Catches the zero-rows trap: successful queries run under the wrong role, account, or filtered view
- Reports verified / not-yet-landed / contradicted as three separate counts, so a partial batch never reads as fully submitted
- Ships with a pre-report checklist and the anti-pattern list this discipline exists to kill
Who it's for: anyone running agents that submit forms, publish content, update CRMs, file records, place orders, or manage batches in external systems. What the buyer gets: one self-contained SKILL.md playbook — workflow, probing rules, reporting format, hard holds, and checklist. Installs into any agent that reads SKILL.md skills; no tools or accounts required.
Requires read access to the target system's state. Cannot verify completion on "fire-and-forget" systems that offer no lookup or census mechanism.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 8 days ago