v0.1.0

Changelog

• c8a7aae fix(release): drop signs block; notarize published macOS archive in Makefile
• c3b6d3c fix(release): notarize signs step produces no .sig (output: false)
• e1f503d feat(dist): npm package + Linux/Windows release builds
• ddc0ffa feat: --version flag (conformance) + OpenCoven PR draft
• d202178 feat(print): system-prompt append, permission-mode, and OpenCoven manifest (SP3+SP4)
• 48ac9fa feat(print): session persistence with --session-id and --resume (OpenCoven SP2)
• 87d765f feat(print): Claude-Code-compatible stream-json protocol (OpenCoven SP1)
• 98ac6c4 docs: add a real README and CONTRIBUTING guide
• 2261894 feat(startup): show a fortune, version, and recent changes under the gopher
• 76e38a3 fix(release): auto-source GITHUB_TOKEN from gh and require signing vars
• d0d643c chore(gitignore): never commit credentials.md or signing keys
• 57bb852 fix(release): make signing skip cleanly on credential-free snapshots
• 12c3f39 chore: remove scratch cmd/hello hello-world
• b81c27a docs: add backlogs, prompt-system plan, and session notes
• bd5ccc1 feat(config): first-run setup wizard, global config, and no baked-in endpoint
• bb1d8dc feat(release): signed+notarized Homebrew release pipeline and version command
• a00207f feat(tui): stop OSC-11 background-query leak and add gopher banner
• 4749760 chore: ignore the compiled gophermind binary
• 73d9f8d docs(security): queue streaming-cancellation entry for deploy
• a9f66aa feat(stream): polish mid-stream cancellation across llm, agent, and TUI
• fce2394 feat(llm): request/response middleware hooks for the HTTP client
• e475577 feat(llm): optional mutual-TLS and custom-CA for secure internal endpoints
• 3d76201 feat(agent): add JSONL transcript export of full message history
• bfec093 feat(sampling): per-request temperature/top-p override via config and /temp TUI command
• 75b8eba feat(llm): fallback model chain for graceful degradation
• 47cd95c feat(llm): probe model capabilities (context window, tool support, max output)
• fc83a95 feat(llm): response caching by prompt hash with TTL
• d3e3471 feat(llm): automatic retry with bounded exponential backoff
• 330470c feat(config): named provider profiles selectable by --profile
• 418a4da feat(llm): streaming token usage & cost meter
• 14bc566 docs(security): coordination note — cwd done, remaining items owned by live fleet
• dc39f38 docs(security): audit log, report, and staged deploy queue
• 0a49387 fix(safety): close sandbox escapes — symlink containment, search arg injection, deny-list bypass
• 61f8d63 feat(tui): full-screen alt-screen layout with scrollable viewport
• 9281594 chore: add bin/build.sh to build and launch gophermind
• dd4858f fix(cli): guard args[1:] so bare invocation doesn't panic
• 8b25eaa feat(tui): commit finished output to scrollback; teatest e2e
• 4ce4129 feat(tui): launch TUI for interactive mode; TTY guard; drop line REPL
• dafa2a7 feat(tui): Bubble Tea interactive session (model, update, view, run)
• b9f0749 feat(agent): stream prose as token events; add Reset
• 4f549f8 feat(llm): streaming completions with tool-call reassembly
• dbc34b0 docs: TUI implementation plan
• 8c8a1d1 docs: TUI design spec (Claude Code-style interface)
• 4ebfa1d feat: gophermind agentic coding harness baseline
• e4adcc5 chore(small-feature): auto-commit after workflow-template
• da30e25 docs: create roadmap (9 phases)
• 3715313 docs: define v1 requirements
• 7a0fbff docs: complete project research
• e6fceaa chore: add project config
• ce89b0d docs: initialize project
• ccaf136 first commit