provider.Request.ResponseFormat was being silently dropped by the
openai translation layer (translate.go:translateRequest). The
upstream provider type and the openai-go SDK both supported it; the
adapter just never propagated it.
This is why Move 1 (set ResponseFormat=ResponseJSON in the SLM
classifier) produced zero observable change: the field made it from
the classifier into provider.Request but stopped at the OpenAI
translation step. The ollama backend (used via the OpenAI-compatible
endpoint) therefore never received format=json_object and kept
emitting free-form prose, which the classifier's downstream JSON
parser duly rejected — 50 fallbacks in a row across two model swaps.
Translate provider.ResponseJSON to oai.ResponseFormatJSONObjectParam
and provider.ResponseText to oai.ResponseFormatTextParam; leave the
union zero-valued when the caller didn't set ResponseFormat so the
SDK omits the field per its omitzero tag. Three table cases cover
the json / text / unset paths.
Affects ollama, llama.cpp, llamafile, and any other backend reached
via openaicompat — all run through openai.translateRequest.
Closes R-4 and R-5 of the routing-defaults plan.
R-4: Strengths + CostWeight defaults for closed frontier models.
Cloud entries land in the same knownFamilyDefaults table as local
ones, with MaxComplexity intentionally left zero (cloud arms get
no complexity ceiling). CostWeight tuned per the plan's rationale:
claude-opus-4-7 → Planning/SecurityReview/Debug/Refactor, 0.3
claude-sonnet-4-6 → Generation/Refactor/Review, 0.7
gpt-5.5 → Planning/SecurityReview/Generation, 0.3
gpt-5.3-codex → Generation/Refactor/Debug/UnitTest, 0.6
gpt-5.2 → Orchestration/Review, 0.8
gemini-3.1-pro → Planning/Review/Orchestration, 0.5
gemini-3.5-flash → Boilerplate/Explain/Orchestration, 1.2
The 0.3 weight on frontier arms keeps them competitive on
SecurityReview / Planning despite $4+/Mtok; 1.2 on Gemini Flash
penalizes cost more so it only wins when cost is genuinely
decisive (boilerplate, explain).
Mechanism: extracted applyFamilyDefaults into defaults.go and call
it from Router.RegisterArm. Single source of truth — both local
discovery and the primary-provider path in cmd/gnoma/main.go now
flow through the same defaults application. Removed the duplicate
apply block from RegisterDiscoveredModels.
Legacy model IDs (claude-opus-4-20250514, gpt-4o, o3, gemini-2.5-pro,
etc.) intentionally do not match any table entry — keeps users on
pinned older models safe from imposed 2026 Strengths.
R-5: gpt-5.3-codex registration.
- internal/provider/openai/provider.go: added to fallbackModels
and inferOpenAIModelCapabilities (400K context, 32K output).
- internal/provider/ratelimits.go: gpt-5.3-codex and its dated
alias gpt-5.3-codex-2026-02-15 added with the same Tier 1
quotas as gpt-5.2.
Gemini 3.x (3.1-pro-preview, 3.5-flash, 3.1-flash-lite) was already
registered in both google/provider.go and ratelimits.go — no change
needed for that part of R-5.
Test coverage:
- ResolveFamilyDefaults table-driven across all 7 cloud entries
including prefix-sharing (gpt-5.5-pro → gpt-5.5 defaults,
gemini-3.1-pro-preview → gemini-3.1-pro defaults).
- Legacy IDs return !ok.
- RegisterArm applies cloud defaults end-to-end.
- User-supplied Strengths and CostWeight are not overridden.
- ID.Model() fallback works when ModelName is empty (test code
often constructs arms this way).
Refs: docs/superpowers/plans/2026-05-23-routing-defaults-refresh.md
When the user message has at least one ImageContent block, build a
ChatCompletionContentPartUnionParam array with text + image_url
parts instead of the string content path. Image bytes are inlined
as a base64 data URL (data:<media-type>;base64,...). Adjacent text
blocks are merged into a single TextContentPart. Pure-text user
messages stay on the existing string fast path.
This covers OpenAI direct + every openaicompat backend (Ollama,
llama.cpp, llamafile) since they all share the same provider.
Tests: pure text uses OfString; image present emits 2 content parts
(text + image_url with the expected base64 payload); nil-Image
blocks are dropped and adjacent text merges correctly.
Bump hard-coded provider defaults to the May 2026 lineup:
- Anthropic: claude-sonnet-4-6 (default); Opus 4.7 and Haiku 4.5 in
the fallback list. 4.6/4.7 generation has 1M context standard.
- OpenAI: gpt-5.5 (default); 5.5-pro / 5.2 / 5.2-chat-latest in
fallback. ThinkingModes now baseline on GPT-5.x.
- Google: gemini-3.5-flash (default); 3.1 Pro / Flash Lite in fallback.
- Mistral: mistral-large-latest unchanged (Mistral Large 3); add
mistral-medium-3.5, mistral-medium-2511, mistral-large-2512 to the
rate-limit map.
Legacy dated IDs retained in fallback lists and ratelimits maps so
configs pinned to claude-sonnet-4-20250514 / gpt-4o / gemini-2.5-flash
keep resolving. Capability tables (ContextWindow, MaxOutput,
ThinkingModes) updated to match each generation. CLI help text in
cmd/gnoma/main.go also updated.
Apply gofmt -w across the codebase (struct field comment realignment
only — no semantic changes) and silence two errcheck warnings on
fmt.Sscanf / fmt.Fprintf return values in internal/router/discovery
with explicit `_, _ =` discards. Required so `make check` is green
before tagging v0.1.0.
Local-model servers (Ollama, llama.cpp, llamafile) routed through the
OpenAI-compatible path frequently emit tool-call arguments that are
*almost* valid JSON — wrapped in markdown fences, padded with prose, or
trailing a stray comma. Strict parsing fails, the engine receives empty
args, and the agent loop has to retry or escalate.
Adds repairArgs(raw) at the EventToolCallDone boundary: strict-parse
first, then apply cheap lexical fixes (strip ```json fences, drop
trailing commas before }/], extract the first balanced {...} block with
proper string/escape awareness). On success, the repaired bytes flow
through unchanged; on failure, the original is returned and downstream
parsing surfaces the error as before.
Frontier providers (OpenAI proper, Anthropic, Mistral, Google) are
unaffected — their SDKs return structured args that pass strict parse.
The repair only does work when the upstream output is malformed.
11 unit tests cover: valid passthrough, empty, trailing commas,
single/double-line fences, prose-wrapped, braces-inside-strings,
multiple top-level objects (takes the first), and unrepairable input.
A stream-level test verifies the wiring through flushNextToolCall.
- OpenAI provider: use Models.ListAutoPaging() to discover available models
- Anthropic provider: use Models.ListAutoPaging() to discover available models
- Google provider: use Models.All() iterator to discover available models
- All providers fall back to hardcoded lists if API calls fail
- Add capability inference functions for each provider based on model ID
- Add tests for model discovery fallback behavior
This enables gnoma to dynamically discover new models as they become available
from cloud providers, while maintaining backward compatibility with fallback
lists for offline use or API failures.
Add EffortLevel (auto/low/medium/high) as a provider-agnostic reasoning
control, replacing the Capabilities.Thinking bool. Each provider maps
the level to its native parameter: Anthropic budget tokens (1K/8K/16K),
OpenAI reasoning_effort (low/medium/high), Google thinking budget
(1K/8K/16K). Task classification auto-infers effort from TaskType and
complexity; filterFeasible excludes arms that lack the required level.
Three compounding bugs prevented tool calling with llama.cpp:
- Stream parser set argsComplete on partial JSON (e.g. "{"), dropping
subsequent argument deltas — fix: use json.Valid to detect completeness
- Missing tool_choice default — llama.cpp needs explicit "auto" to
activate its GBNF grammar constraint; now set when tools are present
- Tool names in history used internal format (fs.ls) while definitions
used API format (fs_ls) — now re-sanitized in translateMessage
Additional changes:
- Disable SDK retries for local providers (500s are deterministic)
- Dynamic capability probing via /props (llama.cpp) and /api/show
(Ollama), replacing hardcoded model prefix list
- Engine respects forced arm ToolUse capability when router is active
- Bundled /init skill with Go template blocks, context-aware for local
vs cloud models, deduplication rules against CLAUDE.md
- Tool result compaction for local models — previous round results
replaced with size markers to stay within small context windows
- Text-only fallback when tool-parse errors occur on local models
- "text-only" TUI indicator when model lacks tool support
- Session ResetError for retry after stream failures
- AllowedTools per-turn filtering in engine buildRequest
Stop retrying llama.cpp 500s that are deterministic tool-parse failures
by inspecting the error message body (ClassifyHTTPError). Wrap OpenAI SDK
errors as ProviderError so the engine's retry logic classifies them. Add
localInitPrompt for local models that uses sequential fs_* calls instead
of spawn_elfs (which local models can't produce reliably).
provider/openai:
- Fix doubled tool call args (argsComplete flag): Ollama sends complete
args in the first streaming chunk then repeats them as delta, causing
doubled JSON and 400 errors in elfs
- Handle fs: prefix (gemma4 uses fs:grep instead of fs.grep)
- Add Reasoning field support for Ollama thinking output
cmd/gnoma:
- Early TTY detection so logger is created with correct destination
before any component gets a reference to it (fixes slog WARN bleed
into TUI textarea)
permission:
- Exempt spawn_elfs and agent tools from safety scanner: elf prompt
text may legitimately mention .env/.ssh/credentials patterns and
should not be blocked
tui/app:
- /init retry chain: no-tool-calls → spawn_elfs nudge → write nudge
(ask for plain text output) → TUI fallback write from streamBuf
- looksLikeAgentsMD + extractMarkdownDoc: validate and clean fallback
content before writing (reject refusals, strip narrative preambles)
- Collapse thinking output to 3 lines; ctrl+o to expand (live stream
and committed messages)
- Stream-level filter for model pseudo-tool-call blocks: suppresses
<<tool_code>>...</tool_code>> and <<function_call>>...<tool_call|>
from entering streamBuf across chunk boundaries
- sanitizeAssistantText regex covers both block formats
- Reset streamFilterClose at every turn start
Thin wrapper over OpenAI adapter with custom base URLs.
Ollama: localhost:11434/v1, llama.cpp: localhost:8080/v1.
No API key required for local providers.
Fixed: initial tool call args captured on first chunk
(Ollama sends complete args in one chunk, not as deltas).
Live verified: text + tool calling with qwen3:14b on Ollama.
Five providers now live: Mistral, Anthropic, OpenAI, Google, Ollama.
Streaming, tool use (index-based delta accumulation), tool name
sanitization (fs.read → fs_read), StreamOptions.IncludeUsage for
token tracking. Hardcoded model list (gpt-4o, gpt-4o-mini, o3, o3-mini).
Wired into CLI with OPENAI_API_KEY env support.
Live verified: text streaming + tool calling with gpt-4o.
vikingowl