- Home
- Skills
- Technical Documentation
- Spec to Checks
Spec to Checks
Plan features and read-only reviews with traceable requirements, tasks and acceptance checks before implementation.
$15
Works with the AI Agents you already use
Spec to Checks
Example session with this skill installed
Plan only: export the currently filtered products as CSV with columns sku, name and price only. Exclude internal_notes. Preserve commas, quotes and newlines. If no rows match, emit a header-only CSV. No repository or test framework is available. Do not implement or execute anything.
- Read your context and instructions
- Compiled the spec to checks
Recorded development example using the 0.3.0 root-only instructions, before the current documentation update. Supporting files were not supplied; this was not a live Agensi connector test. No code or tests were executed. Full observed response:
Goal / exclusions
Plan a CSV export of the currently filtered products, with only sku,name,price. Exclude internal_notes and all other fields. No implementation or execution.
Target / evidence
Observed — user request: preserve commas, quotes and newlines; export headers even when no rows match. No repository or test framework is available.
Proposed: preserve filtered row order and existing field values. Paths and commands are unverified; content pin
PENDING.
Requirements
- R1: Export exactly the currently filtered rows.
- R2: Emit columns
sku,name,price, in that order, with no additional fields. - R3: Preserve commas, quotes and newlines through CSV escaping.
- R4: When no rows match, emit only the header.
Checks
| ID | Scope / inputs | Method | Expected output | Failure signal | Status |
|---|---|---|---|---|---|
| C1 → R1 | Products with matching and excluded rows | Parse export and compare rows against the active filter | Exactly matching rows, in preserved order | Missing, extra or reordered rows | NOT RUN |
| C2 → R2 | Products including internal_notes | Inspect parsed header and field counts | Header exactly sku,name,price; three fields per row | Extra column or leaked notes | NOT RUN |
| C3 → R3 | Values containing commas, ", LF and CRLF | Round-trip through a CSV parser | Original values recovered exactly; quoted fields use doubled internal quotes | Split records, altered values or parse failure | NOT RUN |
| C4 → R4 | Filter matching zero rows | Inspect and parse export | One header record, zero data records | Empty file or data records | NOT RUN |
Tasks / traceability
T1 — R1: Identify the filtered row source within the future export scope. Done when C1 inputs and expected rows are defined.
2.
T2 — R2, R3: Specify field projection and CSV serialization. Done when C2–C3 fixtures and expected results are defined.
3. T3 — R4: Specify the zero-match output. Done when C4 expects a header-only CSV.
Risks / approval
Assumed: existing application rules determine price formatting and null values; those rules remain unspecified without repository context. Verification tooling must be selected separately.
Approve this planning scope? Approval does not authorize implementation or execution.
Connects securely to your tools. The creator never sees your data.
About this skill
Spec to Checks is a planning workflow for your AI agent. It is designed to organize a request into goals, exclusions, numbered requirements, concrete acceptance checks and ordered tasks, with traceability between them.
The instructions tell the agent to separate observations from assumptions, mark proposed checks NOT RUN, ask one scope-changing question when needed, and stop for approval instead of implementing or auditing. Review planning is instructed to distinguish the selected version from an exact content pin and to ask neutral questions rather than invent findings.
Includes the main workflow, fallback plan headings, a reusable template, three fictional worked examples, a quickstart and the license notice. Template/example content may be used when your host supplies it.
Example requests:
- Plan CSV export of filtered products, using sku, name and price only. Do not build it.
- Plan a read-only review of README.md and package.json for my stable release. No fixes or execution.
Use through Agensi's supported agent connection and your own model access. Repository-aware planning requires authorized host-native file reading. No Edu or Moragent installation is needed.
This skill does not implement code, audit code, execute tests or orchestrate agents. It is instructions, not a security sandbox or a guarantee of correctness. Review generated plans before approving them. Approval does not automatically execute the plan or authorize out-of-scope fixes.
Personal, non-transferable use by the individual buyer, including their own professional work; no redistribution of the skill. Eduardo Moraga retains ownership. Agensi's platform refund terms apply; no additional seller refund conditions.
How to use
Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed 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 scanned
Verified clean 1 day ago
- Passed all security checks, Safe to install
Requires