interactiv model groq picker

Author: itsflaid

AGENTS.md

@@ -75,3 +75,14 @@ working on DevMap.
 Update the relevant document whenever implementation or verification changes
 its contents. Do not duplicate these personal notes into the public `docs/`
 folder unless the information is intended for users or contributors.
+
+<!-- DevMap Instruction Block -->
+## DevMap Context
+
+Before working in this repository, read `DEVMAP.md` first.
+Read `.devmap/index.json` first, then the relevant
+`.devmap/features/*.json` map. Inspect files from `sourcePriority` before
+exploring broadly. Use `.devmap/snapshot.json` only when those lightweight
+navigation files are insufficient. If the navigation files are missing, run
+`devmap analyze`.
+<!-- End DevMap Instruction Block -->

DEVMAP.md

@@ -0,0 +1,86 @@
+# DevMap
+
+This repository uses DevMap to create reusable project context for developers and AI agents.
+
+## Project Context
+
+- Detected framework: astro
+- Agent navigation index: `.devmap/index.json`
+- Feature maps: `.devmap/features/*.json`
+- Full analysis archive: `.devmap/snapshot.json`
+- DevMap config: `~/.devmap/config.json`
+
+## Recommended Workflow
+
+1. Run `devmap analyze` after cloning the repository.
+2. Run `devmap analyze --fresh` after major architecture changes.
+3. Use `devmap ask "<question>"` before manually exploring unrelated files.
+4. Treat DevMap output as an architecture overview, then verify critical behavior in source code.
+
+## Commands
+
+```bash
+devmap analyze
+devmap analyze --json
+devmap ask "how does authentication work?"
+devmap ask "where is authentication handled?" --json
+devmap doctor
+devmap doctor --json
+```
+
+## Agent Navigation Contract
+
+This repository uses DevMap as the primary navigation source. Use the lightweight
+navigation files before broad repository exploration.
+
+Preferred reading order:
+
+1. Read `.devmap/index.json`.
+2. Pick the relevant feature using its name and keywords.
+3. Open the matching `.devmap/features/*.json` map.
+4. Inspect only the files listed in `sourcePriority` first.
+5. Read `.devmap/snapshot.json` only when the index and feature maps are
+   insufficient or full archive/debug context is required.
+
+Do not scan the whole repository first.
+
+Open source files only when:
+
+- the snapshot is missing;
+- the snapshot is stale;
+- the snapshot does not contain enough detail;
+- exact implementation is required;
+- the task is edit, debug, or refactor;
+- the user explicitly asks for code changes.
+
+When source inspection is needed, inspect the smallest relevant set first.
+Prefer feature entry points and flow steps over broad folder exploration.
+
+## Required Agent Workflow
+
+1. Read `DEVMAP.md`.
+2. Read `.devmap/index.json`.
+3. Open the relevant feature map.
+4. Inspect at most the smallest relevant source-file set from `sourcePriority`.
+5. Explain which navigation entry guided the decision when giving advice.
+6. Avoid unrelated files unless the navigation data is incomplete or exact
+   code verification is required.
+
+If `.devmap/index.json` or `.devmap/snapshot.json` is missing, run
+`devmap analyze` when DevMap is available and configured. If analyze fails
+because DevMap is not initialized, ask the user to run `devmap init` and then
+`devmap analyze`.
+
+If the snapshot may be stale, run `devmap analyze --fresh` before relying on
+it.
+
+Use `--json` when calling DevMap programmatically so stdout remains one
+parseable JSON document without ANSI or terminal decoration.
+
+Do not edit generated files inside `.devmap/`.
+
+## Repository Safety
+
+- `.devmap/` is local generated state and should stay out of Git.
+- Never commit API keys or provider credentials.
+- DevMap helps locate relevant code; it does not replace source-level verification.

PRD.md

@@ -327,16 +327,11 @@ Run static analysis, generate project snapshot, and output a readable project ov
 
 ```bash
 devmap analyze
-devmap analyze --deep
 ```
 
-**`--deep` flag:**
-
-| | Standard | `--deep` |
-|---|---|---|
-| Analysis scope | Project-level overview | Per-module detailed explanation |
-| AI usage | Lower | Higher |
-| Best for | Quick mapping | Large/unfamiliar projects |
+`devmap analyze` uses the provider and model stored in
+`~/.devmap/config.json`. Groq users choose a model during setup and can change
+it later with `devmap config model <model-id>`.
 
 **Generated files:**
 
@@ -576,9 +571,7 @@ ai/
 | Command | Model | Reason |
 |---|---|---|
 | `analyze` | `openai/gpt-oss-20b` | Balanced architecture interpretation |
-| `analyze --deep` | `openai/gpt-oss-120b` | Heavy cross-module reasoning |
 | `analyze` fallbacks | `qwen/qwen3.6-27b` -> `llama-3.3-70b-versatile` -> `llama-3.1-8b-instant` | Keep snapshot enrichment available across model-specific limits |
-| `analyze --deep` fallbacks | `llama-3.3-70b-versatile` -> `qwen/qwen3.6-27b` -> `openai/gpt-oss-20b` | Degrade heavy reasoning gradually instead of failing immediately |
 
 DevMap retries a rate-limited model up to three times, then advances through
 the command-specific chain. It also advances when a model is unavailable or
@@ -591,6 +584,10 @@ Model availability changes over time. Before changing the default routing,
 verify the current Groq model list and lifecycle status. Preview models must
 not be used as a primary default for a public DevMap release.
 
+Groq setup lists the currently available Groq models after validating the API
+key. Users choose one with arrow keys and Enter; the selected model is stored
+as the user's preferred model.
+
 OpenRouter setup asks for a model ID after validating the API key. Pressing
 Enter selects `openrouter/free`; entering another free or paid model stores
 that exact model as the user's preferred model. Explicit user selections take
@@ -1017,7 +1014,7 @@ Rate limit reached.
 
 DevMap will retry in 12 seconds.
 
-Tip: Use standard analyze instead of --deep for lower token usage.
+Tip: Choose a smaller model with devmap config model <model-id> for lower token usage.
 ```
 
 All error messages must be:

README.md

@@ -253,10 +253,11 @@ AI features require a provider API key. DevMap uses Groq by default — analysis
 | OpenAI     | Planned |
 | Gemini     | Planned |
 
-`devmap init` lets you choose Groq or OpenRouter with the arrow keys. For
-OpenRouter, pressing Enter at `OpenRouter model [openrouter/free]:` keeps the
-free router; typing another model ID uses that free or paid model instead.
-Change it later with `devmap config model <model-id>`.
+`devmap init` lets you choose Groq or OpenRouter with the arrow keys. Groq
+shows the available model list and stores the selected model. For OpenRouter,
+pressing Enter at `OpenRouter model [openrouter/free]:` keeps the free router;
+typing another model ID uses that free or paid model instead. Change it later
+with `devmap config model <model-id>`.
 
 API keys are stored locally:
 

docs/architecture.md

@@ -643,7 +643,6 @@ MVP default model routing:
 | Command          | Primary                    | Ordered fallbacks |
 | ---------------- | -------------------------- | ----------------- |
 | `analyze`        | `openai/gpt-oss-20b`       | `qwen/qwen3.6-27b` -> `llama-3.3-70b-versatile` -> `llama-3.1-8b-instant` |
-| `analyze --deep` | `openai/gpt-oss-120b`      | `llama-3.3-70b-versatile` -> `qwen/qwen3.6-27b` -> `openai/gpt-oss-20b` |
 
 Each model receives up to three exponential-backoff retries for HTTP 429.
 After those retries, or when a model is unavailable or returns HTTP 5xx,
@@ -657,6 +656,10 @@ on 2026-06-20. Recheck provider lifecycle status before publishing a release.
 Users can override automatic routing with `devmap config model <model>`.
 Running `devmap config model auto` restores the defaults above.
 
+Groq setup lists available Groq models after API-key validation and stores the
+selected model in the global config. The stored model is used as the primary
+choice until the user changes it with `devmap config model`.
+
 OpenRouter does not use the Groq command-routing table. During `devmap init`,
 Enter accepts `openrouter/free`, while a typed model ID is stored and used
 exactly as the user selected it. Explicit OpenRouter model choices receive no

docs/commands.md

@@ -44,6 +44,7 @@ devmap init
 * Confirm AI provider
 * Input API key or read environment variable
 * Validate API key
+* Ask for a Groq model from the provider model list
 * Ask for an OpenRouter model; Enter defaults to `openrouter/free`
 * Save global configuration to `~/.devmap/config.json`
 * Detect current project framework
@@ -120,17 +121,12 @@ Analyze the current project and generate a reusable project snapshot.
 
 ### Usage
 
-```bash
-devmap analyze
-devmap analyze --deep
-```
-
-### Modes
-
-| Mode     | Purpose                                |
-| -------- | -------------------------------------- |
-| Standard | Fast project overview                  |
-| `--deep` | More detailed architecture explanation |
+```bash
+devmap analyze
+```
+
+`devmap analyze` uses the model stored in `~/.devmap/config.json`. Change the
+stored model with `devmap config model <model-id>`.
 
 ### Responsibilities
 
@@ -219,33 +215,7 @@ Snapshot saved:
 
 ```
 
-
-### Deep Output
-
-When using:
-
-```bash
-devmap analyze --deep
-```
-
-DevMap adds a deeper module-level explanation.
-
-Example:
-
-```txt
-Module Breakdown
-
-app/
-Main application routes and layouts.
-
-app/api/
-Server-side API routes.
-
-lib/
-Shared utilities, database access, authentication logic, and helpers.
-```
-
-### Rules
+### Rules
 
 * Static analysis must run before AI interpretation
 * Do not send the entire project source to AI
@@ -415,7 +385,10 @@ devmap config model auto
 `auto` restores command-based routing:
 
 * `analyze` uses `openai/gpt-oss-20b`
-* `analyze --deep` uses `openai/gpt-oss-120b`
+
+For Groq, `devmap init` lists available models after API-key validation. Pick a
+model with the arrow keys and press Enter. The selected model is stored in the
+global config.
 
 For OpenRouter, `devmap init` prompts with:
 
@@ -430,7 +403,6 @@ The typed model is stored as the primary choice and is not silently replaced.
 Automatic routing also uses ordered fallback chains:
 
 * `analyze`: `qwen/qwen3.6-27b`, `llama-3.3-70b-versatile`, then `llama-3.1-8b-instant`
-* `analyze --deep`: `llama-3.3-70b-versatile`, `qwen/qwen3.6-27b`, then `openai/gpt-oss-20b`
 
 DevMap advances after model-unavailable and transient provider responses. For
 rate limits, it first retries the current model three times with exponential
@@ -476,7 +448,6 @@ integration.
 ```bash
 devmap init --json
 devmap analyze --json
-devmap analyze --deep --json
 devmap onboarding --json
 devmap doctor --json
 devmap config model auto --json

docs/for-me-personal/PROGRESS.md

@@ -4,6 +4,24 @@ Terakhir diperbarui: 2026-06-23
 
 ## Update 2026-06-23
 
+### Groq Model Picker And Analyze Deep Removal
+
+- `~/.devmap/config.json` tetap menyimpan `provider`, `apiKey`, dan `model`.
+  Config lama tanpa `provider` atau `model` dibaca sebagai `provider: "groq"`
+  dan `model: "auto"` agar tetap backward compatible.
+- `devmap init` untuk Groq sekarang mengambil daftar model dari endpoint model
+  Groq setelah API key valid, lalu menampilkan picker arrow-key/Enter.
+- Model Groq yang dipilih disimpan ke global config dan dapat diganti dengan
+  `devmap config model <model-id>`.
+- OpenRouter setup tidak diubah: tetap memakai prompt model text dengan default
+  `openrouter/free`.
+- Flag `devmap analyze --deep` dihapus dari CLI dan dokumentasi aktif. Hasil
+  investigasi: analyzer static, snapshot, dan cache path-nya sama; perbedaan
+  sebelumnya hanya routing model/fallback, prompt/token limit, dan output
+  `Module Breakdown` kecil.
+
+---
+
 ### Ask Command — Complete Removal
 
 - **Seluruh fitur `devmap ask` dihapus permanen.**

docs/for-me-personal/TEST.md

@@ -52,6 +52,25 @@ Remove-Item Env:OPENROUTER_API_KEY
 Jangan simpan atau menyalin API key nyata ke repository, snapshot, output test,
 atau dokumentasi debugging.
 
+## Groq Model Picker And Analyze Command
+
+Focused automated tests:
+
+```powershell
+pnpm --filter devmap exec tsx --test test/init-and-errors.test.ts test/config-command.test.ts test/analyze-ai.test.ts
+```
+
+Expected:
+
+- Config lama di `~/.devmap/config.json` yang hanya berisi `apiKey` tetap
+  terbaca sebagai Groq dengan `model: "auto"`.
+- Setup Groq interaktif menampilkan daftar model Groq setelah API key valid.
+- Pilih model dengan arrow key lalu Enter; model tersimpan di global config.
+- `devmap config model <model-id>` tetap dapat mengganti model yang tersimpan.
+- `devmap analyze --help` tidak menampilkan flag `--deep`.
+- OpenRouter setup tetap memakai prompt model text dengan default
+  `openrouter/free`.
+
 ## Mixed Workspace Snapshot Accuracy
 
 Jalankan static analyze pada root DevMap dengan config AI terisolasi:
@@ -258,12 +277,8 @@ pnpm --filter devmap exec tsx --test test/config-command.test.ts test/analyze-ai
 
 Expected automatic routing:
 
-- `ask`: `llama-3.1-8b-instant`
 - `analyze`: `openai/gpt-oss-20b`
-- `analyze --deep`: `openai/gpt-oss-120b`
-- `ask` fallback: Qwen 3.6 27B -> Llama 70B Versatile -> GPT-OSS 20B
 - `analyze` fallback: Qwen 3.6 27B -> Llama 70B Versatile -> Llama 8B Instant
-- deep fallback: Llama 70B Versatile -> Qwen 3.6 27B -> GPT-OSS 20B
 
 Automated expectations:
 
@@ -925,12 +940,6 @@ Pastikan:
 - raw provider error dan stack trace tidak muncul;
 - `doctor` menyatakan key dan model valid.
 
-Tes deep model:
-
-```powershell
-devmap analyze --deep --fresh
-```
-
 Hapus key:
 
 ```powershell

packages/cli/README.md

@@ -90,17 +90,15 @@ DevMap-owned server.
 ```bash
 devmap init
 devmap analyze
-devmap analyze --deep
 devmap analyze --fresh
 devmap onboarding --write
 devmap onboarding --write --language id
 devmap doctor
 devmap config model auto
 ```
 
-Groq automatic routing uses a fast model for focused questions and larger
-models for architecture analysis. OpenRouter uses the model selected during
-init. Change either provider's model with:
+Groq and OpenRouter use the model selected during init. Change either
+provider's model with:
 
 ```bash
 devmap config model <model-id>