[Extension]: Update AEE dependency metadata (existing v1.0.0 entry)

Bắt đầu bằng cách xác định mục AEE hiện có trong catalog từ issue #4565 và xem schema của catalog hoặc hướng dẫn validation. Chỉ cập nhật metadata của requires.tools để bổ sung các prerequisite AEE và spec-kit-evaluator đã được sửa, đồng thời giữ nguyên version, URL và các field khác hiện có; khi đó mục catalog sẽ mô tả chính xác archive v1.0.0 đã được publish.

Extension ID

aee

Extension Name

Applied Epistemic Engineering

Version

1.0.0

Description

Challenges claims, traces evidence, propagates uncertainty, and routes epistemic failures between Spec Kit phases.

Author

ElectroHire

Repository URL

https://github.com/electrohire/spec-kit-aee

Download URL

https://github.com/electrohire/spec-kit-aee/archive/refs/tags/v1.0.0.zip

License

MIT

Homepage (optional)

https://github.com/electrohire/spec-kit-aee

Documentation URL (optional)

https://github.com/electrohire/spec-kit-aee#readme

Changelog URL (optional)

https://github.com/electrohire/spec-kit-aee/blob/main/CHANGELOG.md

Required Spec Kit Version

=1.0.0

Required Tools (optional)

• Python >=3.11 — required to run the adapter and engine.
• aee>=1.0.2,<2 — required CLI, installed from the applied-epistemic-engineering Python package. Repository: https://github.com/electrohire/applied-epistemic-engineering. Version 1.0.2 is the actual minimum for the extension's gaps command.
• spec-kit-evaluator>=1.0.0,<2 — required Spec Kit extension, installed with extension ID evaluator, not a shell executable. Repository: https://github.com/electrohire/spec-kit-evaluator. Supplies result contract, compose/report/route commands. Explicit release: https://github.com/electrohire/spec-kit-evaluator/archive/refs/tags/v1.0.0.zip.

These are informational tool/dependency entries; the current extension installer does not auto-install them. Install engine and Evaluator first.

Number of Commands

6

Number of Hooks (optional)

5

Tags

epistemics, evidence, provenance, uncertainty, governance

Key Features

• Explicit stable-ID claim assessment and bounded recovery.
• Declared dependency/conflict graphs and retained-record ledger verification.
• Evaluator Contract results and optional lifecycle hooks.
• Gap coverage register from a verification matrix and recorded test results.

Testing Checklist

• Extension installs successfully via download URL.
• All six adapter operations execute with expected outcomes using engine 1.0.2 (unsupported assessment/gate/challenge intentionally return 1, not 0).
• Documentation is complete and accurate in the unchanged v1.0.0 archive: corrections are in the linked PR; this submission does not pretend that an unreleased patch is already shipped.
• No open CodeQL alerts were returned at audit time; this is a scoped static-scan observation, not a security certification.
• Tested on a real initialized Spec Kit project and the linked benchmark project; benchmark failures remain disclosed.

Submission Requirements

• Existing release includes a valid extension.yml accepted by Spec Kit.
• README.md includes installation and usage instructions; linked PR improves known gaps.
• MIT LICENSE included.
• Existing v1.0.0 GitHub release/tag remains published and unchanged.
• Six command files exist and render; cross-extension placeholders resolve.
• Extension ID is valid.

Testing Details

Windows 11, Python 3.12.6, Spec Kit 1.0.8, AEE engine 1.0.2, and Evaluator 1.0.0 at e6e86e1ca2c99c167375953ad8d682913f3de244. The existing public AEE archive was downloaded without authentication and installed into a clean initialized project, first from its extracted archive and then directly with specify extension add aee --from <published-url>. Evaluator was also installed from its published URL. SHA-256: 7be7e11549ae7f9ebfee740af7f61d5b16098f68d89cc454ce51d0134cc18399. Six commands, five hooks, rendered placeholders, assessment, challenge, graph, ledger verification, gate, and gap output were checked. No paid inference was used.

The source fix adds hash-locked real-engine tests and Linux/Windows installation CI: https://github.com/electrohire/spec-kit-aee/pull/6. Consult its final check statuses. The existing release's manifest incorrectly advertises an engine minimum of 1.0.0 and uses an unenforced requires.commands field. This catalog correction makes the actual prerequisites discoverable now; it does not modify the existing archive or assert that the pending 1.0.1 release exists.

Example Usage

# From an initialized Spec Kit project, with the Python environment activated:
python -m pip install "applied-epistemic-engineering>=1.0.2,<2"
specify extension add evaluator --from https://github.com/electrohire/spec-kit-evaluator/archive/refs/tags/v1.0.0.zip
specify extension add aee --from https://github.com/electrohire/spec-kit-aee/archive/refs/tags/v1.0.0.zip

Then invoke /speckit.aee.assess artifact=claims.json phase=after_specify in the coding agent, using its rendered command spelling.

Proposed Catalog Entry

{
  "aee": {
    "name": "Applied Epistemic Engineering",
    "id": "aee",
    "description": "Challenges claims, traces evidence, propagates uncertainty, and routes epistemic failures between Spec Kit phases.",
    "author": "ElectroHire",
    "version": "1.0.0",
    "download_url": "https://github.com/electrohire/spec-kit-aee/archive/refs/tags/v1.0.0.zip",
    "repository": "https://github.com/electrohire/spec-kit-aee",
    "homepage": "https://github.com/electrohire/spec-kit-aee",
    "documentation": "https://github.com/electrohire/spec-kit-aee#readme",
    "changelog": "https://github.com/electrohire/spec-kit-aee/blob/main/CHANGELOG.md",
    "license": "MIT",
    "category": "process",
    "effect": "read-write",
    "requires": {
      "speckit_version": ">=1.0.0",
      "tools": [
        {
          "name": "python",
          "version": ">=3.11",
          "required": true
        },
        {
          "name": "aee",
          "version": ">=1.0.2,<2",
          "required": true
        },
        {
          "name": "spec-kit-evaluator",
          "version": ">=1.0.0,<2",
          "required": true
        }
      ]
    },
    "provides": {
      "commands": 6,
      "hooks": 5
    },
    "tags": [
      "epistemics",
      "evidence",
      "provenance",
      "uncertainty",
      "governance"
    ],
    "verified": false,
    "downloads": 0,
    "stars": 0,
    "created_at": "2026-09-15T00:00:00Z",
    "updated_at": "2026-09-15T00:00:00Z"
  }
}

Only requires.tools needs correction; retain existing version, download URL and other metadata. Maintainers may apply their normal updated_at timestamp. spec-kit-evaluator denotes the separate extension dependency, not an executable to probe on PATH.

Additional Context

This updates the existing entry from #4565 using the official Extension Submission process; it is not a new extension and is not a request to alter the catalog via PR. The original issue mentioned Evaluator in prose, but the catalog's tools list omitted it. The companion benchmark is https://github.com/electrohire/spec-kit-aee-benchmark; its published results do not establish a general correctness or cost-saving advantage. The source PR is kept separate from this metadata-only catalog correction and remains unmerged. AI assistance was used for audit, edits, testing and submission preparation; evidence and limitations are linked for maintainer review.