Skip to main content

llm-integration

Add or modify LLM/integration providers in integration-api with caller factory wiring, unified provider routing, streaming behavior, and metric/audit compatibility.

Jump to install

Source facts

Repository
rapidaai/voice-ai
Last source activity
August 23, 2026 at 07:23
Detected SKILL.md language
English
Stars
733
Forks
120

Install options

The review-first prompt is selected by default. You can switch to a direct command or download a local copy.

Review the source files

Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.

File Explorer
4 files

Showing SKILL.md

SKILL.md
Source instructions · Read-only preview
name
llm-integration
description
Add or modify LLM/integration providers in integration-api with caller factory wiring, unified provider routing, streaming behavior, and metric/audit compatibility.
# LLM Integration Skill ## Mission Add provider callers that support chat/stream/verify (and optional embedding/reranking) with consistent metrics and unified routing. ## Inputs expected from user 1. Provider capabilities required: chat, stream, verify, embeddings, reranking. 2. Auth style and endpoint shape. 3. Any model parameter or tool-calling constraints. If user does not answer: - Implement chat + stream + verify first. - Add embeddings/reranking only when provider supports them. ## Hard boundaries In scope: - `api/integration-api/internal/caller/<provider>/...` - `api/integration-api/internal/caller/caller.go` - `api/integration-api/internal/type/callers.go` when interface compatibility changes are required - `api/integration-api/api/unified_provider.go` - `pkg/clients/integration/integration_client.go` if routing/client behavior needs updates - provider UI model metadata in `ui/src/providers/<provider>/...` Out of scope: - assistant-api telephony/STT/TTS/EOS/VAD internals ## Implementation workflow 1. Create provider caller package with `llm.go`, `verify-credential.go`, and optional embedding/reranking files. 2. Register provider in `caller.go` factory methods. 3. Ensure unified gRPC route usage supports provider name casing/validation. 4. Emit metrics via `MetricBuilder` semantics (`TIME_TAKEN`, `STATUS`, request ID). 5. Add tests for chat success/failure, streaming, verify credential, optional embedding/reranking. 6. Update UI provider catalogs (`text-models.json` / `models.json`) as needed. ## Governed lifecycle - Classify work as Fast, Standard, or Governed using `DEVELOPMENT_PROCESS.md`; use the full gated lifecycle only for Governed work. - This skill operates only in its assigned phase and path ownership; it may not approve its own plan or code review. - Governed implementation starts only from a coordinator-attested approved plan; Fast and Standard work follows its lighter documented lifecycle. - Return changed-file and verification evidence to the coordinator, then route the complete verified diff to the read-only `code-reviewer`. ## Validation commands - `go test ./api/integration-api/internal/caller/...` - `go test ./api/integration-api/api/...` - `go test ./pkg/clients/integration/...` - `./.claude/skills/llm-integration/scripts/validate.sh --check-diff --provider <provider>`
View on GitHub