Fantasy v0.21.0 natively includes gpt-5.5 and other newer models in
its responsesModelIDs/responsesReasoningModelIDs lists, making our
workaround unnecessary.
- Delete responses_models.go (go:linkname hack + RegisterResponsesModels)
- Delete responses_models_test.go
- Replace isResponsesAPIModel/isResponsesReasoningModel heuristics with
direct openai.IsResponsesModel/openai.IsResponsesReasoningModel calls
- Remove RegisterResponsesModels calls from registry init/reload
- Remove hack documentation from AGENTS.md
- Update all deps (fantasy v0.21.0, smithy-go, ultraviolet, etc.)
Fantasy's hardcoded responsesModelIDs list gates whether a model uses
the Responses API or Chat Completions code path. When a new model
(e.g. gpt-5.5) is added via `kit update-models` but fantasy hasn't
been updated yet, the type mismatch between *ResponsesProviderOptions
and *ProviderOptions causes a crash.
- Add isResponsesAPIModel()/isResponsesReasoningModel() helpers that
supplement fantasy's checks with prefix-based heuristics for modern
OpenAI model families (gpt-4.1+, gpt-5+, o-series, codex, chatgpt)
- Add RegisterResponsesModels() using go:linkname to append missing
model IDs from our database into fantasy's internal slices at init
time and after ReloadGlobalRegistry()
- Replace all direct openai.IsResponsesModel/IsResponsesReasoningModel
calls in providers.go with the new helpers
- Merge embedded + cached model databases instead of cache-only fallback
- Bump fantasy v0.19.0 -> v0.20.0 to match existing import usage
- Document the technique and model-family update process in AGENTS.md
- Remove top-level old_text/new_text params from edit tool schema
- Make edits array the sole interface; single edits pass 1-item array
- Simplify normalizeEditInput, removing dual-mode branching logic
- Update UI renderer to only read from edits array
- Remove old_text/new_text from bodyKeys in message summarizer
- Update web session HTML to iterate edits array
- Convert all single-edit tests to use Edits array
- Replace mixed-mode test with empty-array validation test
Tool blocking via OnToolCall and SetActiveTools returned both a
ToolResponse (IsError=true) and a Go error. Fantasy treats a non-nil
Go error from tool.Run() as a critical failure, aborting the agent
loop without delivering the tool result to the LLM. The model never
saw the block reason and would retry or hallucinate.
- Return nil error for blocked tools (OnToolCall Block=true)
- Return nil error for disabled tools (SetActiveTools)
- Return nil error for extension tool execution failures
- Update tests to assert nil error (IsError response conveys the error)
Fixes#20
- Set context tokens per-step in recordStepUsage instead of waiting
for turn completion; each step re-sends the full conversation so
the reported usage monotonically increases
- Add UsageUpdatedEvent to trigger a TUI re-render after each step
so the status bar reflects updated tokens, cost, and context %
even during gaps between streaming chunks
- Update test to expect per-step context token updates
- Add heightCache map to ScrollList, keyed by item ID, avoiding
repeated Render() calls purely to count lines
- Rewrite GotoBottom() to walk backwards from the end in O(visible)
instead of two full O(N) forward passes over all items
- Replace all height-only Render() calls in clampOffset(), AtBottom(),
ScrollBy(), and ScrollPercent() with cached itemHeight() lookups
- Invalidate cache on width changes (SetWidth) and item mutations
(AppendChunk, AppendStdout/Stderr via InvalidateItemHeight)
- Refresh cache entries in View() from authoritative renders
- Add LLMToolResultOutputContentMedia alias (closes gap in tool result types)
- Add LLMToolResultContentType enum and constants (Text, Error, Media)
- Add LLMToolInfo, LLMProviderOptions, LLMProviderMetadata, LLMPrompt aliases
- Replace all fantasy.* references in hooks.go and hooks_test.go with
SDK-owned aliases, removing the charm.land/fantasy import from both
- Fix gofmt alignment in internal/extensions/symbols.go
- Update SDK skill doc with complete LLM type reference
- Infer Type="image" for image/* MIME types and Type="media" for all
other binary content so the downstream framework creates a media
content block instead of silently discarding Data bytes (#17)
- Extract shared toolOutputToResponse() helper to eliminate duplication
- Add ImageResult() and MediaResult() convenience constructors
- Add LLMToolCall and LLMToolResponse type aliases so SDK consumers
can call Tool.Run() without importing the underlying framework
- Add 6 regression tests covering image, media, and text responses
Closes#17
- Buffer session JSONL writes with bufio.Writer, flush at sync points;
ForkToNewSession and AddLLMMessages now batch N entries into ~1 syscall
- Cache lipgloss styles in style.CachedStyles, lazily built and
invalidated on SetTheme; eliminates ~15 NewStyle() calls per frame in
hot render paths (reasoning blocks, spinner, tool headers, margins)
- Cache git ls-files results for @file suggestions with 3s TTL; typing
@filename no longer spawns 3 subprocesses per keystroke
- Use strings.Builder for StreamingMessageItem.content; eliminates O(n²)
string copying during LLM response streaming
- Set Streaming: true in subagent childOpts to prevent
viper.Set("stream", false) from polluting global state
- Without this, concurrent subagents and the parent could read
stale stream=false from viper, causing provider-level issues
(e.g. Anthropic non-streaming timeouts with extended thinking)
- Update all Go dependencies (bubbletea v2.0.6, fantasy v0.19.0,
acp-go-sdk v0.12.0, mcp-go v0.49.0, and transitive deps)
- Replace SetSessionModel with SetSessionConfigOption to match new
acp-go-sdk Agent interface (union type with ValueId/Boolean variants)
- Add ListSessions stub returning empty list (new required method)
- Refresh embedded_models.json from models.dev/api.json
- Update ACP smoke test: add initialize handshake, session/list,
session/set_config_option, session/cancel, and fix update parsing
- remove "You" label and icon from user messages, use borderless content block
- remove input title bar ("Enter your prompt...") and hint line
- increase textarea from 3 to 4 rows with top/bottom margin
- hide input hints permanently for a cleaner UI
- match separator colors (use theme.Border for both startup and input dividers)
- make startup separator full terminal width instead of hardcoded 80
- add /help for help hint and pipe separators to status bar
- add printCustomMessage/RenderCustomMessage for custom alert labels
- render /help output as markdown with "Help" alert label
- add Ctrl+V (paste image) to help message keys section
- fix reasoning text wrapping using ANSI-aware lipgloss.Style.Width
- export HighlightFileTokens for cross-package use
- Add NoOAuth field to MCPServerConfig with JSON/YAML support
- Guard OAuth error handling and transport setup with the new flag
- Prevents failed dynamic client registration on servers like PubMed
that do not support OAuth
- First ctrl+c clears input and arms quit flag with 3s timeout
- Second ctrl+c within timeout window actually quits
- Show '⚠ Press Ctrl+C again to quit' warning after first press
- Empty input no longer quits immediately on single ctrl+c
- Prompt/overlay states: ctrl+c cancels dialog, re-dispatches to
main handler for double-press tracking instead of quitting
- Update placeholder, help text, and tests to match new behavior
Add charset=utf-8 to Content-Type header and use HTML entity
✓ instead of raw Unicode checkmark to prevent garbled
text display in browsers.
Fixes#9
- wrap thinking text in StreamComponent and render.ReasoningBlock
- plumb width through renderer and streaming item paths
- keeps style consistent with user/assistant blocks and avoids cut-off lines
Adds 'none' thinking level to support OpenAI gpt-5.4 models which use
'reasoning_effort: none' instead of 'minimal'. Includes validation and
auto-adjustment when switching models with incompatible levels.
- Add ThinkingNone constant mapping to ReasoningEffortNone
- Add IsValidThinkingLevelForModel() with gpt-5.4 detection
- Add SuggestThinkingLevelFallback() for level migration
- Auto-adjust thinking level on model switch with user notification
- Update all docs to include 'none' in valid levels
Fixes#11
Add --set-default flag to 'kit auth login' to automatically set the
provider's default model after successful authentication. When no Anthropic
credentials exist but OpenAI credentials are detected, error messages
now suggest using OpenAI with the correct --model flag.
Fixes#9
Change Ctrl+C behavior to match other terminal AI tools (claude, codex, pi):
- First Ctrl+C clears the current input when text is present
- Second Ctrl+C (within 3 seconds) quits the application
- Ctrl+C on empty input quits immediately
- 3-second auto-reset timer clears the 'pressed once' state
- Flag also resets after message submission
Updates placeholder text and help message to reflect new behavior.
Fixes#13
Add defensive validation to detect and prevent cycles in the session tree
parent chain that could occur after compaction or file corruption.
- Add tree_validation.go with cycle detection and parent chain validation
- Validate parent chain before appending messages (AppendMessage)
- Validate firstKeptEntryID exists in AppendCompaction
- Add depth limit and cycle detection to buildTreeNode to prevent infinite recursion
- Log diagnostics on session open to detect existing cycles
- Add tests for cycle detection and graceful handling
Reflect the refactor that made MCPAuthHandler an explicit, opt-in
dependency for remote MCP OAuth. Four surfaces updated:
- README.md: new 'MCP OAuth (remote MCP servers)' subsection under the
Go SDK section, outlining the three consumer patterns (nil / CLI /
custom) and linking to the full options docs.
- pkg/kit/README.md: type cheat-sheet now lists MCPAuthHandler,
DefaultMCPAuthHandler, and CLIMCPAuthHandler alongside the existing
MCPTokenStore entries.
- skills/kit-sdk/SKILL.md: Options example annotated with nil-disables-
OAuth semantics; new 'MCP OAuth Authorization' section precedes the
existing token-storage section; re-exported types list expanded.
- www/pages/sdk/options.md: Options fields table gains MCPAuthHandler
row; new top-level 'MCP OAuth Authorization' section with consumer
matrix, CLI/custom/fully-custom code samples, and a warning callout
about the OnAuthURL nil-hang footgun.
Strip user-facing I/O out of the SDK's OAuth surface so library, daemon,
and web-app embedders are not surprised by port binds or browser opens.
- DefaultMCPAuthHandler no longer calls openBrowser; it exposes an
OnAuthURL(serverName, authURL) hook and performs no presentation I/O.
- kit.New no longer auto-constructs a default handler when
Options.MCPAuthHandler is nil. OAuth is opt-in; remote MCP servers
requiring authorization fail with a clear error if no handler is set.
- CLIMCPAuthHandler owns the CLI policy (browser open + stderr prints)
by wiring an OnAuthURL closure on the inner DefaultMCPAuthHandler.
- openBrowser is now unexported and colocated with its sole caller; no
new exported helper is added to the SDK surface.
BREAKING CHANGE: SDK consumers relying on implicit OAuth with a nil
MCPAuthHandler must now pass kit.NewCLIMCPAuthHandler() (or a custom
implementation) explicitly. The kit CLI is unaffected — cmd/root.go
already constructs the handler explicitly.
TestDetectMediaType/.go fails on CI images (Ubuntu mime-support) where
/etc/mime.types registers '.go → text/x-go', because mime.TypeByExtension
reads those files at init. The test intended to exercise the 'unknown
extension falls through to text/plain' branch but used a real extension,
making the assertion environment-dependent.
Replace '.go' with '.kitsyntheticext', an invented extension that no
system MIME database registers. The fallback path is now exercised
deterministically on any host.
The SDK applies Options by calling viper.Set on viper's process-global
store, which means two Kits constructed in the same process are not
isolated from each other: the second New overwrites the first's keys,
and downstream readers (SetModel, GetThinkingLevel, BuildProviderConfig)
observe the most recent value.
- Add a 'Global viper state warning' block to the Options godoc
explaining the leak, the zero-value-does-not-clear gotcha, and
pointing at viper.Reset() as the migration workaround.
- Add a matching warning to the New godoc so consumers discover the
constraint from either entry point.
- Detach the viperInitMu godoc (previously lodged inside New's comment
block) and clarify that the mutex only guards the construction
window, not instance isolation.
- Add a TODO noting the proper fix: refactor to a per-call viper.New()
instance so each Kit owns its own config store.
- InitConfig now installs a viper env key replacer so keys like
"max-tokens" bind to KIT_MAX_TOKENS under AutomaticEnv; previously
hyphenated keys silently missed their documented env overrides.
- Simplify TestNewPreservesIsSetSemantics: with SkipConfig: true no env
bindings are registered, so the os.Getenv guard and upper() helper
were dead weight. Remove both and drop the unused helper.
The SDK last-resort MaxTokens floor is applied in kit.New() when
Options.MaxTokens, KIT_MAX_TOKENS, .kit.yml, and per-model defaults
are all unset. It was 4096 (inherited from the old setSDKDefaults
viper default) while the CLI --max-tokens cobra default is 8192.
Bump the floor to 8192 so SDK and CLI callers start from the same
base value before rightSizeMaxTokens runs, then update README,
skills/kit-sdk/SKILL.md, and www/pages/{configuration,sdk/options}.md
to match.
Previously setSDKDefaults() registered viper.SetDefault for max-tokens,
temperature, top-p, top-k, frequency/presence-penalty, and thinking-level.
viper.SetDefault makes IsSet() return true, which silently suppressed
per-model defaults (ApplyModelSettings) and automatic right-sizing
(rightSizeMaxTokens) for every SDK-created Kit — and for CLI runs too,
since cmd/root.go routes through kit.New. Effective max-tokens for
claude-sonnet-4-5 was pinned at 4096 instead of 32768.
- Drop SetDefault for all IsSet-sensitive keys; keep only model,
system-prompt, stream, num-gpu-layers, main-gpu.
- Apply a 4096 max-tokens floor directly on the *models.ProviderConfig
struct in kit.New() when nothing else resolved a value. Keeps
viper.IsSet("max-tokens") == false so rightSizeMaxTokens and
per-model maxTokens overrides still fire.
- Update Options.MaxTokens / ThinkingLevel godoc to describe the real
precedence chain.
- Strengthen tests: add Temperature subtest; add
TestNewPreservesIsSetSemantics regression covering all seven keys;
split TestNewWithProviderOptions into three subtests including
Options-beats-viper-state and ProviderURL propagation; add
resetViper helper so subtests don't bleed state.
- Document the new SDK fields (MaxTokens, ThinkingLevel, Temperature,
TopP, TopK, FrequencyPenalty, PresencePenalty, ProviderAPIKey,
ProviderURL, TLSSkipVerify) in README, skills/kit-sdk, and the www
configuration / sdk/options / sdk/overview pages, including a
dedicated precedence table.
Adds programmatic overrides on kit.Options for the model/provider knobs
that were previously only reachable through viper.Set() — letting SDK
consumers (web apps, services, embedded agents) configure kit fully
in-code without polluting global viper state or shipping .kit.yml.
Generation parameters:
- MaxTokens int (max output tokens per response)
- ThinkingLevel string (off/low/medium/high)
- Temperature *float32
- TopP *float32
- TopK *int32
- FrequencyPenalty *float32
- PresencePenalty *float32
Sampling params use pointer types so explicit 0 is distinguishable from
unset; nil leaves provider/per-model defaults in place.
Provider configuration:
- ProviderAPIKey string
- ProviderURL string
- TLSSkipVerify bool
Implementation just pushes Options values into viper inside New(),
so all existing downstream code (BuildProviderConfig, SetModel,
modelSettings lookups, runtime model switching) picks them up
uniformly without any new code paths. Tests added for MaxTokens,
ThinkingLevel, and ProviderAPIKey.
- Raise --max-tokens default from 4096 to 8192.
- Auto-raise MaxTokens toward the model's catalog Limit.Output (capped at
32768) when the user hasn't set --max-tokens explicitly and no per-model
modelSettings override applied. Prevents silent 4k/8k truncation on
models that support 32k-262k output.
- Surface FinishReasonLength at turn end: the app now subscribes to
TurnEndEvent and renders a system-message banner explaining the current
cap, the model's known ceiling, and how to raise it. Previously the TUI
swallowed 'length' stops, producing 'ghost' truncations.
- Export FinishReason* constants on pkg/kit (Stop, Length, ToolCalls,
ContentFilter, Error, Other, Unknown) and fix stale comments that used
Anthropic-style strings.
- Add Kit.MaxTokens() and Kit.MaxOutputLimit() SDK accessors, backed by
Agent.GetMaxTokens() which correctly returns 0 for providers that
suppress the param (e.g. Codex OAuth).
- Tests: rightSizeMaxTokens covers 7 paths (cap, raise, preserve,
explicit flag, nil info, zero limit); handleTurnEnd covers length/
non-length/nil-sendFn and the fallback message formatter.
- Docs: update configuration.md, cli/flags.md, and kit-extensions skill
to reflect the new default and behavior.
- UpdateTheme() only refreshed typography styles, leaving spinner
frames rendered with the old theme's colors
- Now calls knightRiderFrames() to rebuild frames with the new
theme's Primary, Muted, VeryMuted, and MutedBorder colors
- Replace hardcoded nameWidth of 15 with dynamic calculation based on
the longest command name in the filtered list
- Prevents truncation of longer names like /feature-request and
/release-tagger that were cut off with ellipsis
- Cap name column to leave at least 20 chars for descriptions
- Add 1 char gap between name and description columns
- Move thinking toggle from ctrl+t to leader chord (ctrl+x t) to avoid
conflicts with tmux/zellij tab mode and terminal new-tab shortcuts
- Change scrollback jump from alt+home/alt+end to ctrl+home/ctrl+end
for better compatibility across SSH and older tmux versions
- Remove ctrl+d as submit alias (enter suffices); avoids EOF convention
confusion and accidental shell disconnects
- Remove ctrl+a from tree selector filter shortcuts to avoid conflict
with the common tmux prefix remap (ctrl+o cycle still reaches all
filter modes)
Add core TUI support for handling sudo password prompts when executing
bash commands that require elevated privileges.
- Detect sudo commands and check if credentials are cached (sudo -n)
- Show modal password prompt with masked input (• characters) when needed
- Pipe password via stdin using sudo -S -p '' (no password in command string)
- Password flows through context callbacks, never stored in session history
- Add PasswordPromptHandler to agent and SDK event system
- Add password prompt overlay to TUI with 🔐 icon and hidden input
- Include tests for sudo command detection and rewriting
The password is never persisted to disk - it only exists in memory
during execution and is piped directly to sudo via stdin.
- Add InProcessServer field to MCPServerConfig (json:"-", never serialized)
- Add "inprocess" transport type to config, validation, and connection pool
- Add createInProcessClient() using mcp-go client.NewInProcessClient()
- Add Kit.AddInProcessMCPServer() convenience method
- Add Options.InProcessMCPServers for init-time registration
- Export MCPServer type alias (= server.MCPServer) in pkg/kit/types.go
- Add 8 tests covering config, pool, tool manager, and edge cases
- Update SDK README, kit-sdk skill, and www docs
Ed Zynda