docs(plans): estudio del pack opcional Canvas Extensions (GitHub Copilot app)#79
Draft
javiarmesto wants to merge 5 commits into
Draft
docs(plans): estudio del pack opcional Canvas Extensions (GitHub Copilot app)#79javiarmesto wants to merge 5 commits into
javiarmesto wants to merge 5 commits into
Conversation
…lot app) Propuesta decision-ready (Tier 3, Extension) para llevar la salida estructurada del pipeline ALDC a "canvases" visibles y dirigibles en la GitHub Copilot app, con el canvas del al-conductor como piloto y el markdown de .github/plans/ como fuente de verdad. Additivo y read-mostly; no toca Core ni debilita gates HITL; respeta las restricciones de Extension Packs y declared-equals-shipped. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01WtP8csuXp5NntatU51iHgn
Contributor
📋 Collection Validation Results✅ Validation passed! |
…o shipped) Spike de referencia para el estudio canvas-extensions-pack.md §4: extension.mjs (ES module) del canvas del pipeline TDD del Conductor para la GitHub Copilot app, con capabilities agent-callable y el markdown de .github/plans/ como fuente de verdad. No declarado en aldc.yaml, no toca Core ni el spec normativo, sin deps -> no afecta conformidad/CI. Incluye fixture y runner standalone (node extension.mjs) que demuestra el read path sin la Copilot app. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01WtP8csuXp5NntatU51iHgn
Contributor
📋 Collection Validation Results✅ Validation passed! |
…e consumo por agentes Companion del estudio canvas-extensions-pack.md: catalogo de 8 canvases mapeados a las etapas de la metodologia ALDC (spec.create, architect, test-plan, conductor, triage/dredd, presales, build/pr-prepare, memory) y evaluacion de consumo por agentes -> in-host (capabilities que llama el Copilot agent) y cross-host (agentes ALDC via el markdown compartido como esquema de interop). Anade agent-consume.mjs que demuestra el consumo programatico de capabilities (read + HITL write) contra el markdown, sin la Copilot app. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01WtP8csuXp5NntatU51iHgn
Contributor
📋 Collection Validation Results✅ Validation passed! |
… (read path) Demuestra que el patron del catalogo generaliza mas alla del Conductor: nuevo lib.mjs (parsers de markdown puros y reutilizables) + canvases C1 Spec Studio y C3 Test-Plan Checklist (solo read), y demo-all.mjs que carga C1+C3+C4 sobre el mismo plan dir. Refactor del canvas del Conductor (extension.mjs) para usar el core compartido. Sigue additivo bajo el prototipo: no toca aldc.yaml ni el spec normativo. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01WtP8csuXp5NntatU51iHgn
Contributor
📋 Collection Validation Results✅ Validation passed! |
…ead + HITL triage) Quinto canvas del catalogo, enlazado al Audit-Report JSON canonico de Dredd (.github/audits/dredd-audit-*.json, schema de agents/dredd.agent.md Step 4): renderiza verdict + counts por severidad + findings (file:line + cita). El triage HITL (accept/dismiss/assign) se anexa a un sidecar <audit>.triage.json para dejar inmutable el reporte validado en CI (TD-01); ese sidecar es el handoff cross-host dredd -> canvas -> al-developer del catalogo 3B. Incluye fixture de audit canonico y demo-audit.mjs. Sigue additivo bajo el prototipo (sin tocar aldc.yaml ni el spec). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01WtP8csuXp5NntatU51iHgn
Contributor
📋 Collection Validation Results✅ Validation passed! |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Qué es
Estudio decision-ready (no es implementación) sobre adoptar las Canvas Extensions de la GitHub Copilot app dentro de ALDC. Nace de la pregunta: "¿qué opinas de añadir algo de esto al framework?"
Documento:
.github/plans/canvas-extensions-pack.md(registro de estudio, comoclaude-plugin-tool-modernization.md— no usa la plantilla de arquitectura AL, que no aplica a un pack de tooling).Recomendación
Sí, pero como pack opcional Tier-3 (Extension), pilotando solo el canvas del
al-conductor, con el markdown de.github/plans/como fuente de verdad.spec.md,test-plan.md,*-phase-N-complete.md) que hoy queda enterrado como markdown estático.claude-plugin/). Ojo: Canvas es la Copilot app (preview), runtime distinto al de la distribución Copilot actual.Decisiones bloqueantes (HITL)
Riesgos contenidos
Churn de la API en preview · nuevo runtime JS (
extension.mjs) frente al least-privilege del framework · coste de paridad entre superficies · deriva a "segunda fuente de verdad". Todos mitigados manteniendo el pack opcional, aislado de Core y con toda escritura yendo al markdown versionado.🤖 Generated with Claude Code
https://claude.ai/code/session_01WtP8csuXp5NntatU51iHgn
Generated by Claude Code