982012c867
2 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
| aee63bf321 |
feat: phase 5 — AI agent integration (OpenAI, Ollama, Anthropic)
Add the AI layer that turns scan results into suggested next actions. Provider model (internal/ai): - Provider interface with Name() and Suggest(ctx, ScanResult) returning []Suggestion (repo_path, action, message, command, priority 0-2) - NewProvider factory resolves the configured provider and enforces that cloud providers have their API key in the configured env var; Ollama needs no key - OpenAIProvider: Chat Completions with response_format json_object - OllamaProvider: local /api/chat with format:json for structured output - AnthropicProvider: Messages API with system prompt and key/version headers; all three fall back to provider-appropriate defaults for base_url and model - Shared client: 60s timeout, 4 MiB response cap, JSON encode/decode, HTTP error surfaces the upstream status and body Prompt design (BuildPrompt): - Renders the full repository status table (name/status/branch/ahead/ behind/changes) plus strict output rules: exact suggestion schema, allowed actions, smallest-safe-step guidance, no invented repositories, return [] when healthy Parsing (parseSuggestions): - Tolerates ```json fences and surrounding prose, clamps priorities to 0-2, caps the result, and rejects replies without a JSON array Suggestion display (presenter.Suggestions): - "AI SUGGESTIONS" table (repository/action/priority/message/command) rendered below the scan table with priority color-coded (red/yellow/ green); empty results say all repositories are healthy Guarded execution (--ai-execute, experimental): - RunConfirmed executes a suggestion's command inside its repository only after explicit per-command y/N confirmation, and only for actions on an allowlist (commit/push/pull/stash/checkout) so LLM output can never run arbitrary shell commands; cancellation aborts remaining suggestions CLI wiring: - scan: AI block after the table when --ai is set and format is not json (JSON streams stay machine-readable); failures degrade to warnings - watch: AI is queried only on the first frame and when something changed since the previous frame, to avoid hammering the provider every interval - New --ai-execute flag and provider validation in config (openai/ollama/ anthropic), included in the config dump Testing: - httptest-based provider tests verifying request shape (model, auth headers, path), response parsing, API error bodies, HTTP failures, and cancellation - Parse tests: plain/fenced/prose replies, empty arrays, garbage, truncated JSON, priority clamping - Execution tests: unsafe actions and empty commands never run, declined confirmations are skipped, confirmed commands execute in the repo dir, failing commands surface errors - Presenter suggestion table and empty-state tests Verified: go build, go vet, go test -race, gofmt clean; end-to-end smoke test against a local fake Ollama server (request shape confirmed, table + suggestions rendered) and the missing-API-key warning path. |
|||
| d52e0715f2 |
feat: phase 2 — CLI commands and configuration resolution
Add the Cobra-based command surface and the flag/env/config-file resolution layer that all commands share. Configuration (internal/config): - Load() resolves settings with the documented precedence flags > env > config file > defaults, via viper: GITFLOW_-prefixed env vars with dot-to-underscore mapping, plus ~/.gitflow.yaml (or $GITFLOW_CONFIG) - RegisterFlags/NewFlagSet own the flag definitions so every command and the tests share a single source of truth - Config/Validate/Dump cover dir, interval, format (table/json/compact), exclude globs, max depth, worker count, and the AI block (enabled, provider, model, api_key_env, base_url); ConfigFile records the loaded path; Dump renders the effective config as human-readable YAML with the interval as a duration string App orchestration (internal/app): - New() validates the configuration at the boundary (fail fast) - ScanOnce() runs discovery then a concurrent status scan, warning on stderr and continuing when discovery is only partially successful (e.g. permission-denied subtrees), and bundles everything into a ScanResult CLI (cmd/gitflow): - root command with scan / config / version subcommands - scan: resolves config, prompts for the parent directory when stdin is a TTY and --dir was not given (per the README), runs a single pass, and renders the result — interim plain/JSON output until phase 3 lands the presenter package - config: prints the effective configuration - version: prints the build version (ldflags-injectable) - signalContext() wires SIGINT/SIGTERM into a cancellable context for graceful shutdown Testing: - config: defaults, flag overrides, env overrides, flag-beats-env precedence, config file loading (including duration and slice values), GITFLOW_CONFIG path override, validation failures, and Dump output - app: config validation on New, end-to-end ScanOnce over a real temp repo, and missing-directory errors Verified: go build, go vet, go test -race, gofmt clean; manual smoke of `gitflow version`, `gitflow config`, and `gitflow scan -d <dir>` against a scratch directory with a dirty repo. |
