v0.0.14

Changelog

• 59b06b3 chore(deps): update dependency @flanksource/icons to v1.0.63

v0.0.13

Changelog

• 9d7723f fix(security): resolve open Dependabot and CodeQL alerts (#21)

v0.0.12

Changelog

• bf94d0c fix(security): resolve CodeQL code-scanning alerts (#19)

v0.0.11

Changelog

• be598f8 build,: Update gitignore to preserve built webapp for go:embed
• cb1dc95 claude: docs gavel todo execution
• f7b17b3 feat(ai): Add model fallback provider and schema validation middleware
• 4b24050 feat(ai): Add support for Claude and Codex Cmux backends
• 662654d feat(ai): add Gemini 3.0 Pro and improve model family filtering
• eecd4dd feat(ai): add release date tracking and model filtering by currency
• c368bda feat(ai): consult the pricing registry for model suggestions
• b399bc7 feat(ai): drive prompt-run through a spec.Workflow generate→verify loop
• 5911f43 feat(ai): forward SchemaJSON through the agent PromptRequest
• ff59248 feat(ai): suggest the correct model name on an unrecognized model
• e327fa2 feat(ai): universal Prompt.Schema via SchemaJSON + prompt-file output schemas
• e1c1d65 feat(api): add model fallbacks and schema strictness validation
• fa146f4 feat(api,fe,db): Add model fallbacks, prompt schema builder, and gavel session store
• b73fcc9 feat(captainconfig): add PromptDefaults config type and improve test assertions
• 6091ebc feat(claude): classify operational stream types, surface pr-link
• 09da051 feat(claude): surface content-bearing system subtypes as rows
• 74cfc3d feat(cli): RunSessionGet returns the unified session model
• 8cd1d44 feat(cli): add Claude cmux and Codex cmux backend options
• 9e14b4b feat(cli): log prompt source detection (file vs id vs inline vs stdin)
• a27c9d6 feat(cli): session summary cache + context/phase-timing instrumentation
• 9fd7359 feat(cli): unify prompt run/render into one command; deprecate ai prompt
• 496b4d8 feat(cli): warn on a mistyped model at prompt render time too
• d9dae74 feat(cmux): structured CLI option structs + schema-driven editor
• 4b1adf1 feat(fe): add home dashboard with live session monitoring
• 08395cc feat(prompt): add Parse and Document.String for non-rendering .prompt round-trip
• d6a9a08 feat(serve): serve the unified session model at /sessions/{id}
• ad53427 feat(session): add Codex support to the unified model
• 5b6ff78 feat(session): add ToUIMessages and ToReplayEntries projections
• 432f19d feat(session): extract and render Codex inline plans from TodoWrite tools
• f6b3803 feat(session): persistent session store (GORM + embedded postgres)
• b36e9fe feat(session): unified session model + history review fixes
• 3a1455e feat(webapp): adopt shared PromptRunEditor in the prompt workbench
• f0573a1 feat(webapp): session detail consumes the unified session.Session model
• 8f39b94 feat(webapp): show session Server-Timing phases in the browser
• a80e7b0 feat: Add claude-cmux and codex-cmux backend options to AI models filter
• 89270f8 feat: Initialize CLI prompt entity and generate CLI interface
• 42d263e feat: add prompt workbench UI and entity management
• af0f149 feat: add session live monitoring with process discovery
• e6f84b6 feat: add structured output support to claude-agent provider
• e13d371 feat: prompt-run SSE emits the unified session.Message model
• 1f6d39c feat: restructure ai request context and logging
• 51b2487 fix(cli): let sessions get accept the positional id
• 118a286 fix(cli): report unhandled stream types via Warn log, not raw stderr
• 2f394bf fix(cmux): confirm paste landed before pressing Enter on first prompt
• 554b77d fix(webapp): align @types/react on 19 so tsc -b passes
• 3f617ac fix: fallback to config defaults for model and budget fields
• 4437347 refactor(agent): reconcile runner into a generic hook model
• b48ce15 refactor(agent): share spec.Workflow→verify-loop construction
• 2e824c8 refactor(agent): worktree as a thin wt CLI wrapper
• 3d45c50 refactor(api): reframe Workflow as hook declarations
• 192cea8 refactor(cli): extract executePromptRequest function and support new backends
• 4010c39 refactor(fe): remove SpecRuntimeEditor component from RuntimeControls
• 7a684d1 refactor(session): relocate history render provider into pkg/session
• 34ff68a refactor: include cmux backends in AllBackends() and simplify documentation
• 4c44cba refactor: refactor Spec to consolidate permissions and context into Setup
• 05b089b refactor: restructure prompt frontend to api and add prompt run streaming
• 71e125e test(cli): make session scope tests robust to shared-tmp project markers
• 0654197 test(cmux): remove obsolete structured-output-rejection test

v0.0.10

Changelog

• cb1dc95 claude: docs gavel todo execution
• f7b17b3 feat(ai): Add model fallback provider and schema validation middleware
• 4b24050 feat(ai): Add support for Claude and Codex Cmux backends
• 662654d feat(ai): add Gemini 3.0 Pro and improve model family filtering
• eecd4dd feat(ai): add release date tracking and model filtering by currency
• c368bda feat(ai): consult the pricing registry for model suggestions
• b399bc7 feat(ai): drive prompt-run through a spec.Workflow generate→verify loop
• 5911f43 feat(ai): forward SchemaJSON through the agent PromptRequest
• ff59248 feat(ai): suggest the correct model name on an unrecognized model
• e327fa2 feat(ai): universal Prompt.Schema via SchemaJSON + prompt-file output schemas
• e1c1d65 feat(api): add model fallbacks and schema strictness validation
• fa146f4 feat(api,fe,db): Add model fallbacks, prompt schema builder, and gavel session store
• b73fcc9 feat(captainconfig): add PromptDefaults config type and improve test assertions
• 6091ebc feat(claude): classify operational stream types, surface pr-link
• 09da051 feat(claude): surface content-bearing system subtypes as rows
• 74cfc3d feat(cli): RunSessionGet returns the unified session model
• 8cd1d44 feat(cli): add Claude cmux and Codex cmux backend options
• 9e14b4b feat(cli): log prompt source detection (file vs id vs inline vs stdin)
• a27c9d6 feat(cli): session summary cache + context/phase-timing instrumentation
• 9fd7359 feat(cli): unify prompt run/render into one command; deprecate ai prompt
• 496b4d8 feat(cli): warn on a mistyped model at prompt render time too
• d9dae74 feat(cmux): structured CLI option structs + schema-driven editor
• 4b1adf1 feat(fe): add home dashboard with live session monitoring
• 08395cc feat(prompt): add Parse and Document.String for non-rendering .prompt round-trip
• d6a9a08 feat(serve): serve the unified session model at /sessions/{id}
• ad53427 feat(session): add Codex support to the unified model
• 5b6ff78 feat(session): add ToUIMessages and ToReplayEntries projections
• 432f19d feat(session): extract and render Codex inline plans from TodoWrite tools
• f6b3803 feat(session): persistent session store (GORM + embedded postgres)
• b36e9fe feat(session): unified session model + history review fixes
• 3a1455e feat(webapp): adopt shared PromptRunEditor in the prompt workbench
• f0573a1 feat(webapp): session detail consumes the unified session.Session model
• 8f39b94 feat(webapp): show session Server-Timing phases in the browser
• a80e7b0 feat: Add claude-cmux and codex-cmux backend options to AI models filter
• 89270f8 feat: Initialize CLI prompt entity and generate CLI interface
• 42d263e feat: add prompt workbench UI and entity management
• af0f149 feat: add session live monitoring with process discovery
• e6f84b6 feat: add structured output support to claude-agent provider
• e13d371 feat: prompt-run SSE emits the unified session.Message model
• 1f6d39c feat: restructure ai request context and logging
• 0183364 fix(cli): commit built webapp/dist so go:embed build passes in CI
• 51b2487 fix(cli): let sessions get accept the positional id
• 118a286 fix(cli): report unhandled stream types via Warn log, not raw stderr
• 2f394bf fix(cmux): confirm paste landed before pressing Enter on first prompt
• 554b77d fix(webapp): align @types/react on 19 so tsc -b passes
• 3f617ac fix: fallback to config defaults for model and budget fields
• 4437347 refactor(agent): reconcile runner into a generic hook model
• b48ce15 refactor(agent): share spec.Workflow→verify-loop construction
• 2e824c8 refactor(agent): worktree as a thin wt CLI wrapper
• 3d45c50 refactor(api): reframe Workflow as hook declarations
• 192cea8 refactor(cli): extract executePromptRequest function and support new backends
• 4010c39 refactor(fe): remove SpecRuntimeEditor component from RuntimeControls
• 7a684d1 refactor(session): relocate history render provider into pkg/session
• 34ff68a refactor: include cmux backends in AllBackends() and simplify documentation
• 4c44cba refactor: refactor Spec to consolidate permissions and context into Setup
• 05b089b refactor: restructure prompt frontend to api and add prompt run streaming
• 71e125e test(cli): make session scope tests robust to shared-tmp project markers
• 0654197 test(cmux): remove obsolete structured-output-rejection test

v0.0.9

Changelog

• fb737e0 build(deps): pin clicky/aichat to pkg/api-based pseudo-version
• 5cacd2e chore(deps): bump flanksource/commons to v1.53.1
• 6048232 feat scope: cli,claude,tools: add plan command to query session exit-plan-mode plans
• cc79f08 feat(ai): add model catalog system with pricing and caching
• 80a0464 feat(cmux): add interactive claude/codex TUI provider with session tailing
• beb5905 feat(prompt): support spec-native frontmatter in .prompt files
• db7a148 feat: add captain serve command with web UI and chat interface
• adfbba7 feat: add iterative agent loop with verifiers, worktree, and commit
• 624221b feat: add multi-session filtering and .prompt file support to AI commands
• 38298c2 fix(ai): resolve model-catalog test ordering and context-window lookup
• d252f0f fix(provider): buffer runCLI output to avoid Wait/pipe-read race
• e4fc6f7 fix(serve): embed dist placeholder so the binary builds without a prebuilt webapp
• b5a8143 refactor(api): move AI runtime contract into pkg/api
• 438a858 refactor(cli): refactor request model field access patterns
• afe8077 refactor: Refactor cost tracking to break out reasoning and cache token billing
• 19a7090 refactor: consolidate domain types into pkg/api and add session browsing
• a7eab9b refactor: refactor prompt handling to use file-based routing instead of inline truncation
• c72e047 refactor: restructure ai.Request and ai.Config to use nested types
• 03be1fd refactor: restructure ai.Request and ai.Config with nested field groups
• a627774 refactor: restructure ai.Request to use nested api types
• 853ab41 style: gofmt HistoryEntry struct tag alignment

v0.0.8

Changelog

• 8119737 feat(ai): update provider factory to return error and add claude-agent backend
• 6a478da feat(ai/agent): add agent runner with plugin lifecycle and verification
• 8f79a49 feat(ai/history): Add extended tool rendering and session event parsing
• 085d4ca feat(ai/provider): add Claude Agent SDK provider for agentic multi-turn interactions
• 54993a5 feat(ai/provider): add Firebase Genkit provider for Anthropic, OpenAI, and Gemini
• 9cd6475 feat(aichat): add AI chat backend with Genkit, clicky tools, and MCP support
• f1fd16b feat(changes,history,info): track nested sub-agent history
• f2c0461 feat(git): add worktree management and git utilities for agent runs
• 71178b1 feat(history,bash,claude): Add text filtering and improve category classification
• 972adc3 feat(jsonrpc): add newline-delimited JSON-RPC 2.0 client implementation
• 2a354c3 feat: add prompt management
• d0fff1a feat: implement bidirectional tool approval via can_use_tool control protocol
• 8df6cb6 fix(ci): bump clicky to v1.21.33 and fix fixture/container tests
• c67aacb fix(lint): replace deprecated reflect.Ptr with reflect.Pointer
• 87cefb5 refactor(ai): migrate API providers to Firebase Genkit and restructure Claude/Codex backends
• 87dd7cd refactor(ai/provider): migrate claude CLI provider to clicky and reorganize claudeagent
• 2f0b1c1 refactor(aichat): remove aichat package and genkit-based chat implementation
• 415074d refactor(cli): migrate to commons HTTP logger for wire logging
• 0fc0b94 refactor: refactor: introduce package-scoped loggers and improve logging consistency across AI modules
• 0543adf style: apply gofmt

v0.0.7

Changelog

• a0d0a61 feat(ai): unify codex live event rendering with history output
• 3ee16e2 refactor(cli,ai,container,sandbox): extract AIRuntimeOptions and improve error handling

v0.0.6

Changelog

• 2f88588 feat(ai,cli): Add streaming support and loop driver for AI providers
• 5bf3fb0 feat(ai/history): Add support for newer codex live schema and unified history display
• 83a50ef feat(claude): Surface stream-json lifecycle events as synthetic tool entries
• 2fc2758 feat(cli): add interactive configure wizard and config file management
• eb81b9c refactor(ai): migrate from static model catalog to live API listings
• 55f9997 refactor(cli): unify streaming and history output rendering with lineRenderer
• 4c08ccd refactor(collections): extract fuzzy matching logic into reusable collections package

v0.0.5

Changelog

• ce86223 refactor: ai fixture benchmarking