refactor(framework): remove provider-specific model pins from portable artifacts
I maintainer di solito rispondono entro 1 giorno
Nessuno ha ancora preso questa issue.
Valutazione
- Difficoltà
- 5/5
- Tempo stimato
- Più di una settimana
- Idoneità per principianti
- 35/100
- Tipo di issue
- Funzionalità
- Chiarezza
- Abbastanza chiara
- Stato di attività
- Attiva
- Ambito
- developer-experience, tooling
Direzione di ricerca
Start by reading the two agent files, tool-paths.md, tools.md, and the related issues #585, #592, and #654 to map existing model and tool behavior. Done means a maintainer-approved design that assigns vendor-neutral tiers to every shipped SDLC step, supports project and user bindings, defines tool fallbacks and notices, and preserves current behavior when unconfigured.
Scritto dal modello di indicizzazione a partire dal testo della issue.
Descrizione
Outcome
Every framework step inherits the active model and reasoning configuration selected by the user or host. Portable AIDD artifacts contain no provider-specific model override.
Why
The concluded spike found seven Anthropic aliases embedded in shared framework content:
opus:checker,audit,review,debug;sonnet:executor,assert,test.
These values are executable overrides in Claude Code, but not a portable contract. Codex's AIDD translator removes them, while other hosts use different identifiers, inheritance rules, and agent-only configuration surfaces. #700 already demonstrated the failure mode: a framework-pinned model unavailable to the user's account can prevent a session from starting.
The decision is therefore host inheritance only:
- AIDD owns the behavior and responsibility of each skill or agent;
- the user and host own concrete model and reasoning choices;
- AIDD does not add advisory workload tiers until a real consumer and user outcome justify them.
Scope
Includes
- Remove
model: opusandmodel: sonnetfrom the two shipped agents and five shipped skills. - Update assertions, fixtures, snapshots, generated distributions, and documentation that treat those pins as framework defaults.
- Keep the translation layer able to process user-authored host model fields where the target supports them; this task removes AIDD defaults, not host capability.
- Make cross-target tests prove the distributed framework adds no provider-specific model override.
Excludes
- Tier or workload-intent metadata;
- project or personal model mappings;
- account model discovery;
- automatic model switching during an SDLC flow;
- changing a user's host configuration;
- cost measurement or savings claims.
Done when
- No production AIDD skill or agent frontmatter contains
model: opusormodel: sonnet. -
checker,executor,assert,audit,review,test, anddebugretain their existing behavior, tools, and invocation semantics apart from model inheritance. - Builds for every supported target contain no model override originating from framework defaults.
- Translator tests still cover supported user-authored model fields without presenting them as AIDD defaults.
- Documentation states that concrete model and reasoning selection belongs to the user and host.
- Existing setup behavior remains unchanged: AIDD neither writes nor replaces the host's selected model.
Completion evidence
- Repository search over production plugin sources returns no
^model: (opus|sonnet)$match. - Framework build and translation test suites pass for every registered target.
- Generated-package assertions show no AIDD-owned model override.
Relations
- Source: concluded investigation in this issue;
- Related: #700, concrete model ownership and account compatibility;
- Related: #585, portable project-policy boundary;
- Parent: none. This is standalone portability cleanup.
Research sources
- Lingua principale
- TypeScript
- Stelle
- 481
- Fork
- 45
- Merge medio
- 10h 31m
- PR unite (30g)
- 98
Preparare l'ambiente
Come iniziare
- Leggi tutta la issue e poi la guida ai contributi del progetto.
- Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
- Fai un fork del repository e lavora su un branch.
- Apri una pull request che faccia riferimento al numero della issue.
Altre issue di ai-driven-dev/framework
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 82/100
ai-driven-dev/framework#952 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 84/100
ai-driven-dev/framework#940 ·
I maintainer di solito rispondono entro 1 giorno
-
refactor(aidd-orchestrator): the check zone says when to stop, and reviews its axes in one roundAperta
Difficoltà 2/5 1-3 ore Idoneità per principianti 76/100
ai-driven-dev/framework#887 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 84/100
ai-driven-dev/framework#873 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 75/100
ai-driven-dev/framework#625 ·
I maintainer di solito rispondono entro 1 giorno
Tutte le issue di ai-driven-dev/framework
Issue simili
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 68/100
Doist/todoist-cli#576 ·
I maintainer di solito rispondono entro 1 giorno
-
Suggestion: document (or optionally add) a cheaper-model config for find-skills on Claude CodeApertafeature
Difficoltà 1/5 Meno di un'ora Idoneità per principianti 72/100
vercel-labs/skills#2370 ·
I maintainer di solito rispondono entro 1 giorno
-
🐛 Bug supabase/cli
Difficoltà 2/5 1-3 ore Idoneità per principianti 84/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 1/5 Meno di un'ora Idoneità per principianti 90/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 78/100
CopilotKit/aimock#491 ·
I maintainer di solito rispondono entro 1 giorno