Files
gcp_solution_architecture_a…/skills/requirements-chain/SKILL.md
Jonathan Boniface e9246828b3 feat: establish the Target architecture as a graph before code
Design is now approved before any infrastructure code is written (ADR-0005).
A new target-architect Specialist produces one graph of Elements,
Relationships and Checks; the diagram is rendered from it, deployment
guidance must account for every Check, and validation gives a verdict per
Check. Requirements, components, products and recommendations carry ids in
data files, and a submission with a gap in the Requirements chain is refused.

Also in this commit:
- Customer cases group Executions, with an intent history per case
- lesson candidates are captured when an Execution ends
- evaluation gains chain, verdict, judged and interactive checks, and
  per-case thresholds
- Terraform providers are cached between checks
- the requirements analyst proposes defaults instead of asking
- the legacy-baseline case and the checks that only read it are removed
- ADR-0006 (Change requests) accepted; ADR-0004 and ADR-0007 proposed
- plans for iterations 4 to 7 under .agents/plans/

Unproven with a real model: no evaluation run has completed on this design.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-07 13:59:09 +01:00

5.9 KiB

name, description, metadata
name description metadata
requirements-chain Defines the data files that carry requirements through to code and validation: requirements, components, products, recommendations, the Target architecture graph, the implementation map and the verdicts. Use when producing or checking any Deliverable of a solution-architecture Execution, to write its data file in the right format and keep every id traceable.
scope
every Deliverable

Requirements chain

Every requirement must be traceable, by id, to the components that serve it, to the Elements and Checks that realise it, to the code that implements them, and to a validation verdict. Each Deliverable attaches one data file that carries its part of that chain. A submission is refused when a link is missing, and the refusal names it.

Ids are short and stable: reuse an id exactly as the earlier Deliverable wrote it, and never renumber.

The files

requirements.json (requirements)

A list. Number functional requirements FR-1, FR-2 and non-functional ones NFR-1, NFR-2.

[
  {"id": "FR-1", "kind": "functional", "text": "Accept signed webhooks over HTTPS."},
  {"id": "NFR-1", "kind": "non_functional", "text": "Raw events are retained for 30 days."}
]

kind is functional or non_functional. Write one requirement per entry, so each can be checked on its own.

components.json (technical decomposition)

A list of components and the requirement ids each serves. Every functional requirement must be served by at least one component.

[{"id": "C-1", "name": "Ingest endpoint", "requirements": ["FR-1"]}]

products.json (product selection)

One entry per component, naming the product selected for it.

[{"component": "C-1", "product": "Cloud Run"}]

recommendations.json (design recommendations)

A list. Each recommendation names its pillar and the components it applies to; use an empty list for one that applies to the whole workload.

[{"id": "DR-1", "pillar": "security", "text": "Encrypt the topic with a customer-managed key.", "components": ["C-2"]}]

Write one configurable decision per recommendation. "Harden the service" cannot become a Check; "disable public ingress on the worker service" can.

target-architecture.json (Target architecture)

{
  "elements": [
    {"id": "E-1", "name": "Ingest service", "component": "C-1", "product": "Cloud Run", "resource_type": "google_cloud_run_v2_service"},
    {"id": "E-2", "name": "Ingest identity", "product": "IAM", "resource_type": "google_service_account"}
  ],
  "relationships": [
    {"id": "R-1", "from": "E-1", "to": "E-2", "kind": "runs as"}
  ],
  "checks": [
    {"id": "K-1", "applies_to": "E-1", "statement": "ingress is INGRESS_TRAFFIC_ALL and roles/run.invoker is granted to allUsers", "source": "FR-1"},
    {"id": "K-2", "applies_to": "E-2", "statement": "the account is attached to E-1 only", "source": "DR-3"}
  ],
  "not_applicable": [
    {"recommendation": "DR-9", "reason": "Single region by requirement NFR-2."}
  ]
}
  • Elements. One per resource to be created. component is the component it realises; leave it out for a supporting Element such as an identity, a key or a network. product for a component's Element is exactly the product selected for that component.
  • Relationships. from and to are Element ids. kind says what the connection is: calls, publishes to, reads from, runs as, is granted access to.
  • Checks. applies_to is an Element or Relationship id. source is the requirement or recommendation id the Check comes from. Every Element has at least one Check.
  • Completeness. Every component has an Element. Every non-functional requirement and every recommendation is the source of at least one Check, unless the recommendation is listed under not_applicable with a reason. Every selected product appears on its component's Element.

Writing a Check

A Check is a statement that is true or false of the infrastructure code. Name the argument and the value.

  • Verifiable: "kms_key_name is set on the topic", "min_instance_count is at least 2", "the subscription has a dead_letter_policy pointing at E-6".
  • Not verifiable: "the service is secure", "follows least privilege", "scales well".

Split a broad recommendation into several Checks, one per setting. A Check that the validator cannot verify from the code is sent back to you.

implementation.json (deployment guidance)

One entry per Check in the Target architecture, saying where the code satisfies it. file is an attached file; address is the resource address and argument.

[{"check": "K-1", "file": "terraform/main.tf", "address": "google_cloud_run_v2_service.ingest.ingress"}]

Implement every Element, Relationship and Check, and nothing the Target architecture does not ask for. When you believe the Target architecture is wrong or incomplete, say so with needs_input; do not quietly build something else.

verdicts.json (validation results)

{
  "dry_run": {"verdict": "pass", "evidence": "terraform validate: Success! The configuration is valid."},
  "checks": [
    {"check": "K-1", "verdict": "pass", "evidence": "main.tf: ingress = \"INGRESS_TRAFFIC_ALL\"; google_cloud_run_v2_service_iam_member.public"},
    {"check": "K-2", "verdict": "fail", "evidence": "google_service_account.ingest is also attached to the worker service"}
  ],
  "proposed_additions": [
    {"statement": "the bucket has a retention_policy of 30 days", "applies_to": "E-5", "reason": "NFR-1 has no Check on retention"}
  ]
}
  • Give every Check exactly one verdict: pass, fail or not_verifiable, with the evidence you read in the code.
  • dry_run.verdict is pass, fail, or not_run when the tool could not run.
  • Judge the code against the Checks and nothing else. Something the Target architecture lacks is a proposed_additions entry, with the requirement or recommendation id it serves in reason. It is never a failed Check.