Skills

aid-update-testing-strategy

Frontmatter

name — aid-update-testing-strategy
description — Revise the project’s quality (C6) documents — the test landscape (levels, coverage expectations, CI lane mapping, known gaps) and the gate policy (what blocks a merge, the thresholds, who may waive one) — plus any previously created outputs you name. Use this skill when the testing-strategy record already exists and a lane or a gate has changed. Requires no design seed: the change you state in the run is a sufficient input. Reads and consumes a testing-strategy seed when one is present in .aid/design/. When a C6 document does not yet exist, routes to /aid-create-testing-strategy. To author or revise test CODE rather than the strategy, use /aid-create-test or /aid-update-test. Produced by the aid-architect agent and independently verified by aid-reviewer (full verify). Allocates a work-NNN folder.
allowed-tools — Read, Glob, Grep, Bash, Write, Edit, Agent
argument-hint — <change> — what to revise in the testing strategy or the gate policy

Definition: canonical/skills/aid-update-testing-strategy/SKILL.md

Flow

flowchart TB
  classDef aidNode color:#fff
  classDef aidEntry fill:#166534,stroke:#14532d,color:#fff
  classDef aidExit fill:#991b1b,stroke:#7f1d1d,color:#fff
  classDef aidDecision fill:#92400e,stroke:#78350f,color:#fff
  classDef aidLoopBack fill:#1e3a8a,stroke:#1e3a8a,color:#fff
  classDef aidStep fill:#1a2035,stroke:#d4a853,color:#f1f5f9
  n1(["INTAKE"])
  n2["UPDATE"]
  n3["VERIFY"]
  n4(["PRESENT<br/>hard stop -- the user decides"])
  n1 --> n2
  n2 --> n3
  n3 -.-> n2
  n3 --> n4
  class n1 aidEntry
  class n2 aidStep
  class n3 aidLoopBack
  class n4 aidExit
  class n1 aidNode
  class n2 aidNode
  class n3 aidNode
  class n4 aidNode

Source fragments

Every node in the chart above, in chart order, with the exact canonical/ text it was derived from.

1 · INTAKE · entry

## State: INTAKE

1. **Require a stated change.** Empty argument -> ask one bootstrapping question ("What should
   change in the testing strategy or the gate policy?") and wait.
2. **Allocate** through the Work Initiation Gate exactly as the contract's Allocation rule
   requires (`canonical/skills/aid-design/SKILL.md` INTAKE step 4 is the shape), with
   `initiator: aid-update-testing-strategy`. `phase` is not driven.
3. **No seed is required.** The change stated in this run is a sufficient input, and this
   skill completes without a seed. **Read and consume one when present** at
   `.aid/design/testing-strategy.md`, carrying its `## Current direction` into the
   destinations and deleting it once realized.
4. **Resolve the destinations by concern.** This skill binds concern **C6** (quality &
   testing) and owns **two** documents. Where a seed supplied a `## Destination`, use it;
   otherwise apply the concern rule here -- resolve C6 against `.aid/settings.yml`
   `knowledge.doc_set`, falling back to the C6 row of
   `canonical/aid/templates/kb-authoring/domain-doc-matrix.md` -- and **confirm the resolution
   with the user before writing**. Never resolve silently.
5. **Read both destinations** whole, and note each one's `source:` field. An absent
   destination -> route to `/aid-create-testing-strategy` for that document and write nothing
   there.
6. **Classify complexity** for the `aid-architect` dispatch (verifier tier >= producer).

Source: canonical/skills/aid-update-testing-strategy/SKILL.md#L38-L58 · full step: canonical/skills/aid-update-testing-strategy/SKILL.md#L38-L60

2 · UPDATE · step

## State: UPDATE

1. **Apply only the change the user named in this run**, to the owned C6 content, per the
   content rules below.
2. **Ask, every run, which derived outputs to update alongside this one.** This step is
   unconditional -- it runs on every invocation, and the answer is **stored nowhere**: no
   frontmatter backlink, no manifest, no registry, and no state carried between runs. The
   question is asked afresh each time.
3. **Write no tracking metadata** into any output this run touches -- no `derived-from`, no
   `source-doc`, no `generated-by`, no `aid-tracked` field, and no skill-attribution line.
4. **`source: generated` refuses.** A registered build script owns that content.

Source: canonical/skills/aid-update-testing-strategy/SKILL.md#L64-L74 · full step: canonical/skills/aid-update-testing-strategy/SKILL.md#L64-L76

3 · VERIFY · loop-back

## State: VERIFY

**Full verify**, as `design-lifecycle.md` defines it. Not clean -> loop to UPDATE; the
3-cycle circuit breaker there escalates to IMPEDIMENT + `lifecycle: Blocked`.

Source: canonical/skills/aid-update-testing-strategy/SKILL.md#L80-L83 · full step: canonical/skills/aid-update-testing-strategy/SKILL.md#L80-L85

4 · PRESENT — hard stop — the user decides · exit · UNSPECIFIED

## State: PRESENT (hard stop -- the user decides)

Set `lifecycle: Paused-Awaiting-Input`. Present the revision, say which C6 half each change
went to, name every output touched, and assert that a consumed seed is gone.

Source: canonical/skills/aid-update-testing-strategy/SKILL.md#L89-L92 · full step: canonical/skills/aid-update-testing-strategy/SKILL.md#L89-L94

Report an issue with this page →