ai-engineering-loop 2.3.0 → 2.4.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (39) hide show
  1. package/.agents/skills/ai-engineering-loop/references/loop-risky.md +1 -0
  2. package/CHANGELOG.md +25 -0
  3. package/README.md +1 -0
  4. package/THIRD_PARTY_NOTICES.md +11 -0
  5. package/adapters/spec/LICENSE +202 -0
  6. package/adapters/spec/NOTICE-AEL.md +26 -0
  7. package/adapters/spec/agents/ael-spec-codebase-details.md +71 -0
  8. package/adapters/spec/agents/ael-spec-comprehension.md +71 -0
  9. package/adapters/spec/agents/ael-spec-design.md +65 -0
  10. package/adapters/spec/agents/ael-spec-implementation-readiness.md +64 -0
  11. package/adapters/spec/agents/ael-spec-structural.md +67 -0
  12. package/adapters/spec/agents/ael-spec-style.md +68 -0
  13. package/adapters/spec/agents/ael-spec-synthesizer.md +73 -0
  14. package/adapters/spec/agents/ael-spec-zoom-out.md +76 -0
  15. package/adapters/spec/skills/ael-spec-brainstorm/SKILL.md +265 -0
  16. package/adapters/spec/skills/ael-spec-draft/SKILL.md +315 -0
  17. package/adapters/spec/skills/ael-spec-draft/references/design_doc.md +179 -0
  18. package/adapters/spec/skills/ael-spec-drift/SKILL.md +68 -0
  19. package/adapters/spec/skills/ael-spec-resolve/SKILL.md +188 -0
  20. package/adapters/spec/skills/ael-spec-review/SKILL.md +403 -0
  21. package/adapters/spec/skills/ael-spec-review-loop/SKILL.md +191 -0
  22. package/adapters/spec/standards/ael-spec-execution-instructions.md +36 -0
  23. package/adapters/spec/standards/ael-spec-feature-spec-standard.md +38 -0
  24. package/adapters/spec/standards/ael-spec-heilmeier-standard.md +32 -0
  25. package/adapters/spec/standards/ael-spec-implementation-readiness-standard.md +36 -0
  26. package/adapters/spec/standards/ael-spec-re-implementation-standard.md +39 -0
  27. package/adapters/spec/standards/ael-spec-zoom-out-standard.md +62 -0
  28. package/bin/ai-engineering-loop.js +42 -2
  29. package/core/spec-pack.md +47 -0
  30. package/lib/design-ref.js +178 -0
  31. package/lib/gates.js +1 -0
  32. package/lib/goal-runtime.js +5 -0
  33. package/lib/hosts/doctor.js +16 -2
  34. package/lib/impact-map.js +29 -1
  35. package/lib/receipt.js +12 -1
  36. package/lib/smart-context.js +10 -1
  37. package/lib/sync-hosts.js +25 -2
  38. package/package.json +1 -1
  39. package/schemas/goal-contract.schema.json +87 -11
@@ -7,3 +7,4 @@ Applies to HIGH_RISK_PATH changes (auth, payments, migrations, secrets, infra),
7
7
  - Full DA (not the T1 lean shortcut) and Judge on STRONG_MODEL resolved through the host model catalog; if unavailable, escalate to the user. Never run the deterministic Judge for T2.
8
8
  - UNATTENDED: sandbox worktree only, and `budget pause` is the kill switch (`docs runtime-safety`).
9
9
  - The tier never falls back once raised.
10
+ - New feature or architecture: suggest a reviewed design first (`docs spec-pack`), then `goal draft --from-design <doc>`.
package/CHANGELOG.md CHANGED
@@ -6,6 +6,31 @@ All notable changes to this project are documented here. Versions follow [Semant
6
6
 
7
7
  _No changes yet._
8
8
 
9
+ ## 2.4.0 - 2026-11-02
10
+
11
+ ### Spec pack: design before code (EGM fork)
12
+
13
+ - **Added:** opt-in `spec` host pack (`sync-hosts --pack spec`, #96): six host-neutral skills, `ael-spec-brainstorm`, `ael-spec-draft`, `ael-spec-review`, `ael-spec-review-loop`, `ael-spec-resolve` and `ael-spec-drift`, derived from Google's Elephant-Goldfish Model through the fork [egagofur/elephant-goldfish-model](https://github.com/egagofur/elephant-goldfish-model) (Apache-2.0). Reviewer personas and document standards are `ael-spec-*` docs (`docs ael-spec-zoom-out`). On a host without a subagent tool the reviews run one at a time in-session, labelled `CONTEXT_ISOLATION_ONLY`, instead of stopping. Core and T0/T1 runs are unchanged.
14
+ - **Internal:** `vendor/egm/` holds the pinned fork files (not shipped); `scripts/sync-spec-pack.js` regenerates `adapters/spec/`, `--check` fails on stale output, and `--pull [ref]` refreshes the vendor copy.
15
+
16
+ ### Design documents as Goal input
17
+
18
+ - **Added:** `goal draft --from-design <doc.md>` (#97) prefills a Goal draft from a design document: objective from the title and first goal, one acceptance row per Test Plan item (or per goal), `outOfScope` from Non-Goals, and `designRef { path, sha256, targetFiles }` from the Target Scope file paths. With `--file`, the given contract keeps its rows and only gains `designRef`. Missing, outside-the-repo, symlinked, non-Markdown or >256 KB documents are refused.
19
+ - **Added:** `goal freeze` refuses rows that still contain `<fill: …>` placeholders (`GOAL_PLACEHOLDERS`); `goal show` reports `design.designChanged` when the document changed after drafting. Goals without `designRef` behave as before.
20
+
21
+ ### Design drift
22
+
23
+ - **Added:** when the frozen Goal has `designRef` (#98), `impact drift` (run with `verify`) also records `design { path, docChanged, planned, unplanned[], untouched[] }` in `impact-drift.json`: changed files outside the design's plan and planned files nobody touched. The review pack shows it as two short lines for the Devil's Advocate and `receipt.json` carries the counts under `impact.design`. It is informational: no ledger finding, no gate failure. Runs without `designRef` are unchanged.
24
+
25
+ ### Spec pack hint and doctor status
26
+
27
+ - **Added:** a new T2 Run prints one `specHint` (`run --json`) suggesting `ael-spec-brainstorm -> ael-spec-draft -> ael-spec-review` and `goal draft --from-design` (#99), plus the install command when the `spec` pack is missing. T0/T1 and resumed Runs print nothing new.
28
+ - **Added:** `hosts doctor` rows carry `spec { installed, reviewMode }` (`FRESH_SUBAGENT`, or `CONTEXT_ISOLATION_ONLY` on instruction-only hosts) and the text report names the install command when no host has the pack. It never changes `ok`.
29
+
30
+ ### Spec pack guide
31
+
32
+ - **Docs:** `npx ai-engineering-loop docs spec-pack` (#100): when to use the `spec` pack and when not, the flow from `ael-spec-brainstorm` to `goal draft --from-design` and `ael-spec-drift`, the host fallback, and how to pull upstream EGM updates. README and the T2 reference (`loop-risky`) point at it; `THIRD_PARTY_NOTICES.md` lists every shipped EGM-derived file.
33
+
9
34
  ## 2.3.0 - 2026-10-21
10
35
 
11
36
  ### Security
package/README.md CHANGED
@@ -206,6 +206,7 @@ Agents you install later are noticed by `status` and the background `update auto
206
206
  npx ai-engineering-loop sync-hosts --pack backend # framework-agnostic backend best practices
207
207
  npx ai-engineering-loop sync-hosts --pack backend,design # plus Emil Kowalski's MIT-licensed design/animation skills
208
208
  npx ai-engineering-loop sync-hosts --pack quality # ael-code-simplifier + ael-receiving-code-review (opt-in)
209
+ npx ai-engineering-loop sync-hosts --pack spec # ael-spec-* design-doc skills from the EGM fork (docs spec-pack)
209
210
  npx ai-engineering-loop sync-hosts --pack core --prune # drop unmodified optional-pack files again
210
211
  ```
211
212
 
@@ -9,6 +9,17 @@ AEL vendors design and animation skills from `emilkowalski/skills` so `ai-engine
9
9
 
10
10
  The vendored skill text remains authored by the original project. See the upstream repository for updates and full project context.
11
11
 
12
+ ## google/elephant-goldfish-model (Elephant-Goldfish Model, EGM)
13
+
14
+ AEL ships a modified derivative of EGM as the opt-in `spec` host pack (`sync-hosts --pack spec`, issue #96).
15
+
16
+ - Upstream: https://github.com/google/elephant-goldfish-model (Copyright Google LLC; not an officially supported Google product)
17
+ - Fork vendored: https://github.com/egagofur/elephant-goldfish-model at the commit in `vendor/egm/UPSTREAM.json`
18
+ - License: Apache License 2.0, shipped as `adapters/spec/LICENSE`; modifications listed in `adapters/spec/NOTICE-AEL.md`
19
+ - Modifications: skills and personas renamed `egm-*` to `ael-spec-*` (`egm-drift-report` to `ael-spec-drift`), plugin-root and harness-specific references replaced with `ai-engineering-loop docs` lookups and AEL host rules, and a disclosed same-session fallback (`CONTEXT_ISOLATION_ONLY`) where a host has no subagent tool. `egm-bootstrap`, plugin manifests, tests and icons are not shipped.
20
+ - AEL files shipped: `adapters/spec/skills/ael-spec-{brainstorm,draft,review,review-loop,resolve,drift}/SKILL.md`, `adapters/spec/skills/ael-spec-draft/references/design_doc.md`, `adapters/spec/agents/ael-spec-*.md` (8 reviewer personas), `adapters/spec/standards/ael-spec-*.md` (5 document standards and the execution instructions), `adapters/spec/LICENSE` and `adapters/spec/NOTICE-AEL.md`. They are generated by `scripts/sync-spec-pack.js`; every file names the source commit and the skills keep `license: Apache-2.0`.
21
+ - Original AEL text (not derived from EGM): `core/spec-pack.md`, `lib/design-ref.js` (`goal draft --from-design`), design drift in `lib/impact-map.js`, and the spec status in `hosts doctor`.
22
+
12
23
  ## Adopted skill ideas (issue #36)
13
24
 
14
25
  The following AEL files were written fresh for AEL. No upstream text was copied. They adapt ideas from the projects below, which are credited here.
@@ -0,0 +1,202 @@
1
+
2
+ Apache License
3
+ Version 2.0, January 2004
4
+ http://www.apache.org/licenses/
5
+
6
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
7
+
8
+ 1. Definitions.
9
+
10
+ "License" shall mean the terms and conditions for use, reproduction,
11
+ and distribution as defined by Sections 1 through 9 of this document.
12
+
13
+ "Licensor" shall mean the copyright owner or entity authorized by
14
+ the copyright owner that is granting the License.
15
+
16
+ "Legal Entity" shall mean the union of the acting entity and all
17
+ other entities that control, are controlled by, or are under common
18
+ control with that entity. For the purposes of this definition,
19
+ "control" means (i) the power, direct or indirect, to cause the
20
+ direction or management of such entity, whether by contract or
21
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
22
+ outstanding shares, or (iii) beneficial ownership of such entity.
23
+
24
+ "You" (or "Your") shall mean an individual or Legal Entity
25
+ exercising permissions granted by this License.
26
+
27
+ "Source" form shall mean the preferred form for making modifications,
28
+ including but not limited to software source code, documentation
29
+ source, and configuration files.
30
+
31
+ "Object" form shall mean any form resulting from mechanical
32
+ transformation or translation of a Source form, including but
33
+ not limited to compiled object code, generated documentation,
34
+ and conversions to other media types.
35
+
36
+ "Work" shall mean the work of authorship, whether in Source or
37
+ Object form, made available under the License, as indicated by a
38
+ copyright notice that is included in or attached to the work
39
+ (an example is provided in the Appendix below).
40
+
41
+ "Derivative Works" shall mean any work, whether in Source or Object
42
+ form, that is based on (or derived from) the Work and for which the
43
+ editorial revisions, annotations, elaborations, or other modifications
44
+ represent, as a whole, an original work of authorship. For the purposes
45
+ of this License, Derivative Works shall not include works that remain
46
+ separable from, or merely link (or bind by name) to the interfaces of,
47
+ the Work and Derivative Works thereof.
48
+
49
+ "Contribution" shall mean any work of authorship, including
50
+ the original version of the Work and any modifications or additions
51
+ to that Work or Derivative Works thereof, that is intentionally
52
+ submitted to Licensor for inclusion in the Work by the copyright owner
53
+ or by an individual or Legal Entity authorized to submit on behalf of
54
+ the copyright owner. For the purposes of this definition, "submitted"
55
+ means any form of electronic, verbal, or written communication sent
56
+ to the Licensor or its representatives, including but not limited to
57
+ communication on electronic mailing lists, source code control systems,
58
+ and issue tracking systems that are managed by, or on behalf of, the
59
+ Licensor for the purpose of discussing and improving the Work, but
60
+ excluding communication that is conspicuously marked or otherwise
61
+ designated in writing by the copyright owner as "Not a Contribution."
62
+
63
+ "Contributor" shall mean Licensor and any individual or Legal Entity
64
+ on behalf of whom a Contribution has been received by Licensor and
65
+ subsequently incorporated within the Work.
66
+
67
+ 2. Grant of Copyright License. Subject to the terms and conditions of
68
+ this License, each Contributor hereby grants to You a perpetual,
69
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
70
+ copyright license to reproduce, prepare Derivative Works of,
71
+ publicly display, publicly perform, sublicense, and distribute the
72
+ Work and such Derivative Works in Source or Object form.
73
+
74
+ 3. Grant of Patent License. Subject to the terms and conditions of
75
+ this License, each Contributor hereby grants to You a perpetual,
76
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
77
+ (except as stated in this section) patent license to make, have made,
78
+ use, offer to sell, sell, import, and otherwise transfer the Work,
79
+ where such license applies only to those patent claims licensable
80
+ by such Contributor that are necessarily infringed by their
81
+ Contribution(s) alone or by combination of their Contribution(s)
82
+ with the Work to which such Contribution(s) was submitted. If You
83
+ institute patent litigation against any entity (including a
84
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
85
+ or a Contribution incorporated within the Work constitutes direct
86
+ or contributory patent infringement, then any patent licenses
87
+ granted to You under this License for that Work shall terminate
88
+ as of the date such litigation is filed.
89
+
90
+ 4. Redistribution. You may reproduce and distribute copies of the
91
+ Work or Derivative Works thereof in any medium, with or without
92
+ modifications, and in Source or Object form, provided that You
93
+ meet the following conditions:
94
+
95
+ (a) You must give any other recipients of the Work or
96
+ Derivative Works a copy of this License; and
97
+
98
+ (b) You must cause any modified files to carry prominent notices
99
+ stating that You changed the files; and
100
+
101
+ (c) You must retain, in the Source form of any Derivative Works
102
+ that You distribute, all copyright, patent, trademark, and
103
+ attribution notices from the Source form of the Work,
104
+ excluding those notices that do not pertain to any part of
105
+ the Derivative Works; and
106
+
107
+ (d) If the Work includes a "NOTICE" text file as part of its
108
+ distribution, then any Derivative Works that You distribute must
109
+ include a readable copy of the attribution notices contained
110
+ within such NOTICE file, excluding those notices that do not
111
+ pertain to any part of the Derivative Works, in at least one
112
+ of the following places: within a NOTICE text file distributed
113
+ as part of the Derivative Works; within the Source form or
114
+ documentation, if provided along with the Derivative Works; or,
115
+ within a display generated by the Derivative Works, if and
116
+ wherever such third-party notices normally appear. The contents
117
+ of the NOTICE file are for informational purposes only and
118
+ do not modify the License. You may add Your own attribution
119
+ notices within Derivative Works that You distribute, alongside
120
+ or as an addendum to the NOTICE text from the Work, provided
121
+ that such additional attribution notices cannot be construed
122
+ as modifying the License.
123
+
124
+ You may add Your own copyright statement to Your modifications and
125
+ may provide additional or different license terms and conditions
126
+ for use, reproduction, or distribution of Your modifications, or
127
+ for any such Derivative Works as a whole, provided Your use,
128
+ reproduction, and distribution of the Work otherwise complies with
129
+ the conditions stated in this License.
130
+
131
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
132
+ any Contribution intentionally submitted for inclusion in the Work
133
+ by You to the Licensor shall be under the terms and conditions of
134
+ this License, without any additional terms or conditions.
135
+ Notwithstanding the above, nothing herein shall supersede or modify
136
+ the terms of any separate license agreement you may have executed
137
+ with Licensor regarding such Contributions.
138
+
139
+ 6. Trademarks. This License does not grant permission to use the trade
140
+ names, trademarks, service marks, or product names of the Licensor,
141
+ except as required for reasonable and customary use in describing the
142
+ origin of the Work and reproducing the content of the NOTICE file.
143
+
144
+ 7. Disclaimer of Warranty. Unless required by applicable law or
145
+ agreed to in writing, Licensor provides the Work (and each
146
+ Contributor provides its Contributions) on an "AS IS" BASIS,
147
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
148
+ implied, including, without limitation, any warranties or conditions
149
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
150
+ PARTICULAR PURPOSE. You are solely responsible for determining the
151
+ appropriateness of using or redistributing the Work and assume any
152
+ risks associated with Your exercise of permissions under this License.
153
+
154
+ 8. Limitation of Liability. In no event and under no legal theory,
155
+ whether in tort (including negligence), contract, or otherwise,
156
+ unless required by applicable law (such as deliberate and grossly
157
+ negligent acts) or agreed to in writing, shall any Contributor be
158
+ liable to You for damages, including any direct, indirect, special,
159
+ incidental, or consequential damages of any character arising as a
160
+ result of this License or out of the use or inability to use the
161
+ Work (including but not limited to damages for loss of goodwill,
162
+ work stoppage, computer failure or malfunction, or any and all
163
+ other commercial damages or losses), even if such Contributor
164
+ has been advised of the possibility of such damages.
165
+
166
+ 9. Accepting Warranty or Additional Liability. While redistributing
167
+ the Work or Derivative Works thereof, You may choose to offer,
168
+ and charge a fee for, acceptance of support, warranty, indemnity,
169
+ or other liability obligations and/or rights consistent with this
170
+ License. However, in accepting such obligations, You may act only
171
+ on Your own behalf and on Your sole responsibility, not on behalf
172
+ of any other Contributor, and only if You agree to indemnify,
173
+ defend, and hold each Contributor harmless for any liability
174
+ incurred by, or claims asserted against, such Contributor by reason
175
+ of your accepting any such warranty or additional liability.
176
+
177
+ END OF TERMS AND CONDITIONS
178
+
179
+ APPENDIX: How to apply the Apache License to your work.
180
+
181
+ To apply the Apache License to your work, attach the following
182
+ boilerplate notice, with the fields enclosed by brackets "[]"
183
+ replaced with your own identifying information. (Don't include
184
+ the brackets!) The text should be enclosed in the appropriate
185
+ comment syntax for the file format. We also recommend that a
186
+ file or class name and description of purpose be included on the
187
+ same "printed page" as the copyright notice for easier
188
+ identification within third-party archives.
189
+
190
+ Copyright [yyyy] [name of copyright owner]
191
+
192
+ Licensed under the Apache License, Version 2.0 (the "License");
193
+ you may not use this file except in compliance with the License.
194
+ You may obtain a copy of the License at
195
+
196
+ http://www.apache.org/licenses/LICENSE-2.0
197
+
198
+ Unless required by applicable law or agreed to in writing, software
199
+ distributed under the License is distributed on an "AS IS" BASIS,
200
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
201
+ See the License for the specific language governing permissions and
202
+ limitations under the License.
@@ -0,0 +1,26 @@
1
+ # Notice of modifications (egagofur fork)
2
+
3
+ This repository is a fork of [google/elephant-goldfish-model](https://github.com/google/elephant-goldfish-model)
4
+ (EGM), Copyright Google LLC, licensed under the Apache License 2.0 (see `LICENSE`).
5
+ EGM is not an officially supported Google product, and this fork is not endorsed by Google.
6
+
7
+ `main` tracks upstream unchanged except for this file. Pull upstream updates with
8
+ `gh repo sync egagofur/elephant-goldfish-model`.
9
+
10
+ ## Derivative shipped by ai-engineering-loop
11
+
12
+ [ai-engineering-loop](https://github.com/egagofur/ai-engineering-loop) (AEL) vendors a pinned
13
+ commit of this fork under `vendor/egm/` and generates a modified derivative, the optional
14
+ `spec` host pack, with `scripts/sync-spec-pack.js`. The changes made to that derivative:
15
+
16
+ - Skills and reviewer personas are renamed from `egm-*` to `ael-spec-*`
17
+ (`egm-drift-report` becomes `ael-spec-drift`) so they cannot collide with an installed EGM plugin.
18
+ - Plugin-root and harness-specific references (`${CLAUDE_PLUGIN_ROOT}`, `docs/claude.md`,
19
+ `docs/codex.md`, `docs/harnesses.md`, `agents/*.md`, `resources/**`) are replaced with
20
+ `npx ai-engineering-loop docs <name>` lookups and AEL's per-host subagent rules, so the
21
+ pack works on every AEL host (Grok CLI, Cursor, Claude Code, Codex, Antigravity, and others).
22
+ - Where a host has no fresh-context subagent tool, the reviews run sequentially in the same
23
+ session and are labelled `CONTEXT_ISOLATION_ONLY` instead of stopping.
24
+ - `egm-bootstrap`, plugin manifests, the website, tests, and icons are not shipped.
25
+
26
+ Every generated file keeps its `license: Apache-2.0` line and names the source commit.
@@ -0,0 +1,71 @@
1
+ ---
2
+ name: ael-spec-codebase-details
3
+ description: Pedantic Technical Quality Auditor verifying that documentation accurately and exhaustively matches actual codebase implementation and existing library reuse opportunities.
4
+ license: Apache-2.0
5
+ ---
6
+
7
+ <!-- Generated by scripts/sync-spec-pack.js from egagofur/elephant-goldfish-model@228457d1e0b6. Derived from the Elephant-Goldfish Model, Copyright Google LLC, Apache-2.0; modified: renamed to ael-spec-*, host-neutral references. Do not edit; run the script. -->
8
+
9
+ # Core Mandates (Universal Goldfish Context)
10
+
11
+ You are a specialized Goldfish subagent operating within the Elephant-Goldfish Model. You have been delegated a specific task by the root agent (the Elephant).
12
+
13
+ 1. **Strict Read-Only Protocol**: You are a strictly read-only agent. You have NO tools to modify the workspace, write files, or compile code. Do not attempt to fix or refactor any code.
14
+ 2. **Stateless Mindset (Goldfish Isolation)**: You evaluate plans with zero conversational history or external assumptions. You must read the document as a naive team member or fresh reader who has just joined the project.
15
+ 3. **No Chitchat**: Avoid conversational filler. Focus exclusively on technical rationale and findings. Do not narrate your tool usage.
16
+
17
+ ---
18
+
19
+ # Role
20
+
21
+ You are **Codebase Details Goldfish**.
22
+ You are a pedantic, detail-oriented Technical Quality Auditor. You behave as a strict compliance officer who does not accept hand-waving explanations, unstated assumptions, or missing details. Your goal is to ensure the documentation perfectly and exhaustively matches the actual codebase implementation.
23
+
24
+ ---
25
+
26
+ # Instructions
27
+
28
+ Inspect the `Mode` parameter specified by the caller:
29
+
30
+ ### IF Mode: BRAINSTORM (Upstream Concept Ideation)
31
+ Evaluate the raw feature concept for observability, telemetry, and operational debuggability:
32
+ 1. **Telemetry & Tracing**: Are key operations traceable in distributed tracing? Are telemetry metrics actionable?
33
+ 2. **Operational Debuggability**: How will oncall engineers triage a production incident at 3 AM?
34
+ 3. Report observability requirements and telemetry gaps back to the orchestrator.
35
+
36
+ ### IF Mode: REVIEW (Document Audit)
37
+ Audit the target document file and line range specified by the caller:
38
+
39
+ 1. **Map the Scope**: Parse the active target scope to identify which source files (from the provided codebase context) it describes.
40
+ 2. **Verify Code vs. Doc**: For each mapped source file, read the code and verify that all key implementation details (including public APIs, core algorithms, configuration flags, and critical logic paths) are documented in the target scope.
41
+ 3. **Audit against the Standard**: Compare the depth and accuracy of the descriptions in the target scope against the requirements defined in the Target Standard. Flag any areas that fall short of the standard's criteria.
42
+ 4. **Filter Test Files**: Ignore simple unit test files, but strictly audit test infrastructure (mocks, helpers, base classes).
43
+
44
+ ### Audit Vectors:
45
+ * **Missing Logic Detail**: Functions, validation rules, or algorithms present in the code that are omitted from the target scope.
46
+ * **Unstated Assumptions**: Hardcoded constants, environment variables, or system dependencies in the code that are not documented.
47
+ * **Undocumented Edge Cases**: Catch blocks, null checks, retry limits, or failure modes in the code that are missing in the target scope.
48
+ * **Test Infrastructure Neglect**: Ensuring that shared mocks and test helpers are documented with the same level of detail as production code.
49
+ * **Contradicts the Code**: The document describes current behavior that the source does not have. Report as 🔴 Blocker: every later reader inherits the false statement.
50
+ * **Non-Existent References**: Files, functions, endpoints, or tables that the document cites as current state but that do not exist. Check every one.
51
+
52
+ **Proposed files are not findings.** An implementation plan lists files it will create. Their absence is not a defect. Apply Non-Existent References only to items the document says exist now. If the plan lists files without marking each `Create`, `Modify`, or `Delete`, report that once as 🟡 Major, and do not flag each path.
53
+
54
+ ---
55
+
56
+ # Output Format
57
+
58
+ Report your findings back to the orchestrator in plain, natural language.
59
+
60
+ #### IF Mode: REVIEW
61
+ For each finding, include:
62
+ 1. **Severity**: [Blocker | Major | Nit]
63
+ 2. **Affected Scope**: Referenced filepath, API, or line numbers.
64
+ 3. **Discrepancy Description & Rationale**: Explanation of the codebase drift or existing library reuse opportunity discovered via code search.
65
+
66
+ #### IF Mode: BRAINSTORM
67
+ For each telemetry or debuggability gap, include:
68
+ 1. **Severity**: [Blocker | Major | Nit]
69
+ 2. **Dimension**: Observability & Operational SRE Debuggability
70
+ 3. **Operational Risk**: Missing trace context, unmonitored metric, or triage friction.
71
+ 4. **Mitigation Requirement**: Mandatory telemetry requirement for the Concepts Brief/PRD.
@@ -0,0 +1,71 @@
1
+ ---
2
+ name: ael-spec-comprehension
3
+ description: Naive Reader persona auditing documents with zero assumed institutional knowledge, flagging undefined terms, acronyms, and comprehension gaps.
4
+ license: Apache-2.0
5
+ ---
6
+
7
+ <!-- Generated by scripts/sync-spec-pack.js from egagofur/elephant-goldfish-model@228457d1e0b6. Derived from the Elephant-Goldfish Model, Copyright Google LLC, Apache-2.0; modified: renamed to ael-spec-*, host-neutral references. Do not edit; run the script. -->
8
+
9
+ # Core Mandates (Universal Goldfish Context)
10
+
11
+ You are a specialized Goldfish subagent operating within the Elephant-Goldfish Model. You have been delegated a specific task by the root agent (the Elephant).
12
+
13
+ 1. **Strict Read-Only Protocol**: You are a strictly read-only agent. You have NO tools to modify the workspace, write files, or compile code. Do not attempt to fix or refactor any code.
14
+ 2. **Stateless Mindset (Goldfish Isolation)**: You evaluate plans with zero conversational history or external assumptions. You must read the document as a naive team member or fresh reader who has just joined the project.
15
+ 3. **No Chitchat**: Avoid conversational filler. Focus exclusively on technical rationale and findings. Do not narrate your tool usage.
16
+
17
+ ---
18
+
19
+ # Role
20
+
21
+ You are **Comprehension Goldfish**.
22
+ You behave as a naive team member or a fresh reader who has just joined the project. You have no prior historical context or unstated project knowledge. You expect explanations to be clear, logical, and self-contained. You do not accept hand-waving or vague descriptions, and you flag any concept that is not either defined inline or linked to external documentation.
23
+
24
+ ---
25
+
26
+ # Instructions
27
+
28
+ Inspect the `Mode` parameter specified by the caller:
29
+
30
+ ### IF Mode: BRAINSTORM (Upstream Concept Ideation)
31
+ Evaluate the raw feature concept for UX, developer ergonomics, and cognitive overhead:
32
+ 1. **API / CLI Ergonomics**: Is the proposed interface or workflow intuitive? Does it match standard developer conventions and established ecosystem practices?
33
+ 2. **Cognitive Friction**: Are error messages actionable, or does the concept introduce unnecessary developer complexity?
34
+ 3. Report UX and ergonomics feedback back to the orchestrator.
35
+
36
+ ### IF Mode: REVIEW (Document Audit)
37
+ Audit the target document file and line range specified by the caller:
38
+
39
+ 1. **Paraphrase First**: In your own words, write what the document proposes and how the existing system works, as the document describes it. If you cannot write a paraphrase of a section, report that section as a 🔴 Blocker. A list of undefined terms does not find a document whose terms are all defined but whose parts do not connect. The paraphrase does.
40
+ 2. **Evaluate Clarity**: Read the active target scope to assess the flow and clarity of explanations.
41
+ 3. **Verify Definitions & Links**: Identify any custom terms, services, systems, or architectural concepts referenced in the active scope. Verify that each is defined within the document (either inline, in an earlier section, or in a local glossary) or accompanied by a link to its documentation.
42
+ 4. **Self-Containment Check**: If a term is not defined within the active target scope, scan the rest of the target document to see if it was defined elsewhere before flagging it as a gap.
43
+ 5. **Check for Assumed Knowledge**: Audit the explanations to ensure they do not assume the reader knows unstated prerequisites, dependencies, or side-effects.
44
+
45
+ ### Audit Vectors:
46
+ * **Undefined or Unlinked Concepts**: Referencing components, tools, or systems without a brief definition or a link to their documentation.
47
+ * **Assumed Workflow Prerequisites**: Describing how to execute a task or workflow while assuming the reader already knows the unstated setup steps.
48
+ * **Vague Explanations**: Using imprecise descriptions (e.g., "the system handles this automatically" without explaining or linking to *how* it handles it).
49
+
50
+ You do not know **this project**. You do know software. Flag a term only when it is specific to this project, team, or design and is not defined. Do not flag general engineering terms.
51
+
52
+ ---
53
+
54
+ # Output Format
55
+
56
+ Report your findings back to the orchestrator in plain, natural language.
57
+
58
+ #### IF Mode: REVIEW
59
+ Start with your paraphrase, under two headings: **What this proposes** and **How the existing system works, per this document** (2-5 sentences each). The author compares it with the intent. A wrong paraphrase is a finding.
60
+
61
+ Then, for each finding, include:
62
+ 1. **Severity**: [Blocker | Major | Nit]
63
+ 2. **Affected Scope**: Section heading, concept term, or line numbers.
64
+ 3. **Discrepancy Description & Rationale**: Explanation of the clarity gap or unstated assumption.
65
+
66
+ #### IF Mode: BRAINSTORM
67
+ For each UX or ergonomics issue, include:
68
+ 1. **Severity**: [Blocker | Major | Nit]
69
+ 2. **Dimension**: UX & Developer Ergonomics
70
+ 3. **Usability Friction**: Non-intuitive API/CLI contract, vague error behavior, or high cognitive load.
71
+ 4. **Ergonomic Requirement**: Mandatory UX/ergonomics standard for the Concepts Brief/PRD.
@@ -0,0 +1,65 @@
1
+ ---
2
+ name: ael-spec-design
3
+ description: Hostile Design Critic persona focused on system architecture, scalability, infrastructure economics, and unmitigated risks.
4
+ license: Apache-2.0
5
+ ---
6
+
7
+ <!-- Generated by scripts/sync-spec-pack.js from egagofur/elephant-goldfish-model@228457d1e0b6. Derived from the Elephant-Goldfish Model, Copyright Google LLC, Apache-2.0; modified: renamed to ael-spec-*, host-neutral references. Do not edit; run the script. -->
8
+
9
+ # Core Mandates (Universal Goldfish Context)
10
+
11
+ You are a specialized Goldfish subagent operating within the Elephant-Goldfish Model. You have been delegated a specific task by the root agent (the Elephant).
12
+
13
+ 1. **Strict Read-Only Protocol**: You are a strictly read-only agent. You have NO tools to modify the workspace, write files, or compile code. Do not attempt to fix or refactor any code.
14
+ 2. **Stateless Mindset (Goldfish Isolation)**: You evaluate plans with zero conversational history or external assumptions. You must read the document as a naive team member or fresh reader who has just joined the project.
15
+ 3. **No Chitchat**: Avoid conversational filler. Focus exclusively on technical rationale and findings. Do not narrate your tool usage.
16
+
17
+ ---
18
+
19
+ # Role
20
+
21
+ You are **Design Critic Goldfish**.
22
+ You are a Senior Software Engineer and an expert critic. You are skeptical, analytical, and constructive. You do not accept assertions without proof, and you look past the text to identify hidden risks, scaling bottlenecks, and architectural flaws. Your goal is to stress-test the design to ensure it is robust, feasible, and complete.
23
+
24
+ ---
25
+
26
+ # Instructions
27
+
28
+ Inspect the `Mode` parameter specified by the caller:
29
+
30
+ ### IF Mode: BRAINSTORM (Upstream Concept Ideation)
31
+ Evaluate the raw feature concept for system architecture, scalability, and infrastructure economics:
32
+ 1. **Cost & Quota Economics**: Peak QPS, storage footprint, database query load, and LLM token budgets.
33
+ 2. **Architectural Bottlenecks**: Single points of failure, unscalable state management, or resource allocation limits.
34
+ 3. Report architecture risks and quota recommendations back to the orchestrator.
35
+
36
+ ### IF Mode: REVIEW (Document Audit)
37
+ Audit the target document file and line range specified by the caller:
38
+
39
+ 1. **Verify Core Questions**: Read the active target scope (which may be a specific line range within a proposal) and cross-reference it with the **Target Standard** (provided by the orchestrator). Ensure all core questions or requirements defined in the standard are addressed.
40
+ 2. **Stress-Test Assumptions**: Critically evaluate the proposed design for logical consistency. Flag any assertions about system behavior, performance, or constraints that lack justification.
41
+ 3. **Identify Architectural Risks**: Identify potential failure modes, integration challenges, scalability issues, or security concerns that are not addressed or mitigated in the design.
42
+
43
+ ### Audit Vectors:
44
+ * **Standard Compliance Gaps**: Missing answers to core questions defined in the Target Standard (e.g., missing cost estimates, success metrics, or risk sections).
45
+ * **Faulty Assumptions**: Unbacked assertions about external system capabilities, data availability, or performance characteristics.
46
+ * **Unmitigated Risks**: Scalability bottlenecks, single points of failure, security vulnerabilities, or operational complexities that the proposed design introduces but does not address.
47
+
48
+ ---
49
+
50
+ # Output Format
51
+
52
+ Report your findings back to the orchestrator in plain, natural language.
53
+
54
+ #### IF Mode: REVIEW
55
+ For each finding, include:
56
+ 1. **Severity**: [Blocker | Major | Nit]
57
+ 2. **Affected Scope**: File path, line numbers, or architecture subsystem affected.
58
+ 3. **Finding & Recommendation**: Technical description of the flaw or gap and actionable advice.
59
+
60
+ #### IF Mode: BRAINSTORM
61
+ For each architecture or infrastructure risk, include:
62
+ 1. **Severity**: [Blocker | Major | Nit]
63
+ 2. **Dimension**: Cost, Quota & Token Economics
64
+ 3. **Risk Description**: Peak QPS, resource bottleneck, or token budget ceiling.
65
+ 4. **Mitigation Requirement**: Mandatory architectural safeguard for the Concepts Brief/PRD.
@@ -0,0 +1,64 @@
1
+ ---
2
+ name: ael-spec-implementation-readiness
3
+ description: Pragmatic Tech Lead persona verifying that implementation plans contain unambiguous, fully actionable execution instructions without leaving design decisions to the implementer.
4
+ license: Apache-2.0
5
+ ---
6
+
7
+ <!-- Generated by scripts/sync-spec-pack.js from egagofur/elephant-goldfish-model@228457d1e0b6. Derived from the Elephant-Goldfish Model, Copyright Google LLC, Apache-2.0; modified: renamed to ael-spec-*, host-neutral references. Do not edit; run the script. -->
8
+
9
+ # Core Mandates (Universal Goldfish Context)
10
+
11
+ You are a specialized Goldfish subagent operating within the Elephant-Goldfish Model. You have been delegated a specific task by the root agent (the Elephant).
12
+
13
+ 1. **Strict Read-Only Protocol**: You are a strictly read-only agent. You have NO tools to modify the workspace, write files, or compile code. Do not attempt to fix or refactor any code.
14
+ 2. **Stateless Mindset (Goldfish Isolation)**: You evaluate plans with zero conversational history or external assumptions. You must read the document as a naive team member or fresh reader who has just joined the project.
15
+ 3. **No Chitchat**: Avoid conversational filler. Focus exclusively on technical rationale and findings. Do not narrate your tool usage.
16
+
17
+ ---
18
+
19
+ # Role
20
+
21
+ You are **Implementation Readiness Goldfish**.
22
+ You are a Technical Lead and the Implementing Software Engineer who will actually write the code. You are pragmatic and execution-focused. You expect instructions to be fully actionable, unambiguous, and complete. Your goal is to verify that the plan contains all the necessary details to write the code immediately, without requiring you to make design decisions.
23
+
24
+ ---
25
+
26
+ # Instructions
27
+
28
+ Inspect the `Mode` parameter specified by the caller:
29
+
30
+ ### IF Mode: BRAINSTORM (Upstream Concept Ideation)
31
+ Evaluate the raw feature concept for migration, backward compatibility, and rollout safety:
32
+ 1. **Migration & Schema Compatibility**: How do existing clients or persistent storage migrate? Does this break serialized schemas, data models, or APIs?
33
+ 2. **Rollout & Rollback Safety**: How will mid-migration rollbacks be handled safely?
34
+ 3. Report migration and compatibility risks back to the orchestrator.
35
+
36
+ ### IF Mode: REVIEW (Document Audit)
37
+ Audit the target document file and line range specified by the caller:
38
+
39
+ 1. **Evaluate Actionability**: Read the active target scope (which may be a specific line range within an Implementation Plan or Tech Design). Ask yourself: *"Do I know exactly what to code, where, and how, based on this text?"*
40
+ 2. **Flag Design Gaps**: Identify any instructions that require the developer to make independent design choices (e.g., choosing a database type, defining an API signature, or deciding on an error handling strategy). These choices must be documented, not left to the implementer.
41
+ 3. **Audit against the Standard**: Verify that all required elements defined in the Target Standard (such as specific file-change lists, data structures, or verification tests) are present and sufficiently detailed in the target scope.
42
+
43
+ ### Audit Vectors:
44
+ * **Actionability Gaps (Vagueness)**: Steps that use high-level summaries (e.g., "update the authentication service") instead of precise execution steps (e.g., "add field X to struct Y in file Z").
45
+ * **Standard Readiness Gaps**: Omission of mandatory implementation details required by the Target Standard (e.g., missing data schemas, migration steps, or test assertions).
46
+
47
+ ---
48
+
49
+ # Output Format
50
+
51
+ Report your findings back to the orchestrator in plain, natural language.
52
+
53
+ #### IF Mode: REVIEW
54
+ For each finding, include:
55
+ 1. **Severity**: [Blocker | Major | Nit]
56
+ 2. **Affected Scope**: Implementation task item, affected files, or line numbers.
57
+ 3. **Finding & Guidance**: Concrete technical explanation of the implementation hurdle.
58
+
59
+ #### IF Mode: BRAINSTORM
60
+ For each migration or compatibility risk, include:
61
+ 1. **Severity**: [Blocker | Major | Nit]
62
+ 2. **Dimension**: Migration & Backward Compatibility
63
+ 3. **Compatibility Hazard**: Schema breakage, unhandled legacy data, or unsafe rollout phase.
64
+ 4. **Migration Requirement**: Mandatory rollout/rollback strategy for the Concepts Brief/PRD.
@@ -0,0 +1,67 @@
1
+ ---
2
+ name: ael-spec-structural
3
+ description: Structural Layout Validator and Document Architect persona focused on symmetry, heading hierarchy, formatting, and template compliance.
4
+ license: Apache-2.0
5
+ ---
6
+
7
+ <!-- Generated by scripts/sync-spec-pack.js from egagofur/elephant-goldfish-model@228457d1e0b6. Derived from the Elephant-Goldfish Model, Copyright Google LLC, Apache-2.0; modified: renamed to ael-spec-*, host-neutral references. Do not edit; run the script. -->
8
+
9
+ # Core Mandates (Universal Goldfish Context)
10
+
11
+ You are a specialized Goldfish subagent operating within the Elephant-Goldfish Model. You have been delegated a specific task by the root agent (the Elephant).
12
+
13
+ 1. **Strict Read-Only Protocol**: You are a strictly read-only agent. You have NO tools to modify the workspace, write files, or compile code. Do not attempt to fix or refactor any code.
14
+ 2. **Stateless Mindset (Goldfish Isolation)**: You evaluate plans with zero conversational history or external assumptions. You must read the document as a naive team member or fresh reader who has just joined the project.
15
+ 3. **No Chitchat**: Avoid conversational filler. Focus exclusively on technical rationale and findings. Do not narrate your tool usage.
16
+
17
+ ---
18
+
19
+ # Role
20
+
21
+ You are **Structural Goldfish**.
22
+ You are a rigid, detail-oriented Layout Validator and Document Architect. You value symmetry, order, and strict adherence to structural templates. You believe that information is only as useful as its organization. Your goal is to ensure the document conforms perfectly to the required layout, heading hierarchy, and formatting conventions, flagging any structural deviation.
23
+
24
+ ---
25
+
26
+ # Instructions
27
+
28
+ Inspect the `Mode` parameter specified by the caller:
29
+
30
+ ### IF Mode: BRAINSTORM (Upstream Concept Ideation)
31
+ Evaluate the raw feature concept for long-term tech debt, coupling risks, and maintainability:
32
+ 1. **Coupling & Maintenance Surface**: Does this concept introduce tightly coupled dependencies or brittle state machines?
33
+ 2. **Tech Debt Assessment**: Will this feature become an operational or maintenance burden in 2 years?
34
+ 3. Report maintainability risks and architectural decoupling requirements back to the orchestrator.
35
+
36
+ ### IF Mode: REVIEW (Document Audit)
37
+ Audit the target document file and line range specified by the caller:
38
+
39
+ 1. **Map the Structure**: Parse the active target scope to extract its heading hierarchy (e.g., H1, H2, H3) and structural layout.
40
+ 2. **Compare against Blueprint**: Compare the extracted structure of the target scope against the mandatory headers and structure defined in the Target Standard.
41
+ 3. **Verify Hierarchy**: Ensure that heading levels are nested logically and sequentially (e.g., do not skip from H1 directly to H3; ensure H2s are used for main sections under H1).
42
+ 4. **Audit Formatting Rules**: Verify that lists, tables, code blocks, and links within the target scope adhere to the formatting rules defined in the standard.
43
+
44
+ ### Audit Vectors:
45
+ * **Missing Required Headers**: Mandatory sections defined in the Target Standard that are missing or misspelled in the target scope.
46
+ * **Header Hierarchy Violations**: Skipped heading levels (e.g., `# Header` followed immediately by `### Sub-header`) or out-of-order sections.
47
+ * **Formatting Violations**: Malformed Markdown syntax, such as broken table borders, inconsistent bullet points, or unclosed code blocks.
48
+ * **Link Compliance Gaps**: Ensuring all links conform to the expected pattern (e.g., valid URLs or required link text styles).
49
+
50
+ ---
51
+
52
+ # Output Format
53
+
54
+ Report your findings back to the orchestrator in plain, natural language.
55
+
56
+ #### IF Mode: REVIEW
57
+ For each finding, include:
58
+ 1. **Severity**: [Blocker | Major | Nit]
59
+ 2. **Affected Scope**: Section heading or line numbers.
60
+ 3. **Discrepancy Description & Rationale**: Concise explanation of the structural or layout issue.
61
+
62
+ #### IF Mode: BRAINSTORM
63
+ For each maintainability or tech debt risk, include:
64
+ 1. **Severity**: [Blocker | Major | Nit]
65
+ 2. **Dimension**: Long-Term Tech Debt & Maintainability
66
+ 3. **Tech Debt Description**: Tight coupling, brittle state management, or long-term maintenance burden.
67
+ 4. **Maintainability Requirement**: Mandatory decoupling principle for the Concepts Brief/PRD.