keyhunter/.planning/phases/03-tier-3-9-providers/03-02-PLAN.md

phase, plan, type, wave, depends_on, files_modified, autonomous, requirements, must_haves

phase	plan	type	wave	depends_on	files_modified	autonomous	requirements	must_haves
03-tier-3-9-providers	02	execute	1		providers/perplexity.yaml providers/you.yaml providers/voyage.yaml providers/jina.yaml providers/unstructured.yaml providers/assemblyai.yaml providers/deepgram.yaml providers/elevenlabs.yaml providers/stability.yaml providers/runway.yaml providers/midjourney.yaml pkg/providers/definitions/perplexity.yaml pkg/providers/definitions/you.yaml pkg/providers/definitions/voyage.yaml pkg/providers/definitions/jina.yaml pkg/providers/definitions/unstructured.yaml pkg/providers/definitions/assemblyai.yaml pkg/providers/definitions/deepgram.yaml pkg/providers/definitions/elevenlabs.yaml pkg/providers/definitions/stability.yaml pkg/providers/definitions/runway.yaml pkg/providers/definitions/midjourney.yaml	true	PROV-03	truths artifacts key_links 11 new Tier 3 Specialized provider YAMLs load (huggingface already exists → 12 total) Providers with documented prefixes (pplx-, xai-voyage, sk-, jina_) use tight regex Voice/image providers without documented prefixes use keyword-only path provides contains providers/perplexity.yaml Perplexity pplx- prefix pattern pplx- path provides contains providers/elevenlabs.yaml ElevenLabs API key keyword detection elevenlabs path provides contains providers/assemblyai.yaml AssemblyAI keyword detection assemblyai from to via pattern provider keywords[] Registry Aho-Corasick automaton NewRegistry() keywords

Create 11 Tier 3 Specialized LLM/AI provider YAMLs — search (Perplexity, You.com), embeddings (Voyage, Jina, Unstructured), voice (AssemblyAI, Deepgram, ElevenLabs), and image/video (Stability, Runway, Midjourney). Huggingface is pre-existing from Phase 2 and must NOT be recreated. Total Tier 3 = 12 after this plan.

Purpose: Satisfy PROV-03 (12 Tier 3 Specialized providers).

Output: 22 YAML files (11 providers × 2 locations).

Addresses PROV-03.

<execution_context> @$HOME/.claude/get-shit-done/workflows/execute-plan.md @$HOME/.claude/get-shit-done/templates/summary.md </execution_context>

@.planning/ROADMAP.md @.planning/phases/03-tier-3-9-providers/03-CONTEXT.md @pkg/providers/schema.go @providers/huggingface.yaml @providers/cohere.yaml Schema: pkg/providers/schema.go. No `category` field. Confidence: high|medium|low. Patterns field may be omitted for keyword-only providers. Keywords ≥3 required. Dual-location required (providers/ + pkg/providers/definitions/). RE2 regex only.

IMPORTANT: providers/huggingface.yaml ALREADY EXISTS (tier: 3). Do NOT modify or recreate it.

Task 1: Search + embeddings providers (Perplexity, You.com, Voyage, Jina, Unstructured, AssemblyAI) providers/perplexity.yaml, providers/you.yaml, providers/voyage.yaml, providers/jina.yaml, providers/unstructured.yaml, providers/assemblyai.yaml, pkg/providers/definitions/perplexity.yaml, pkg/providers/definitions/you.yaml, pkg/providers/definitions/voyage.yaml, pkg/providers/definitions/jina.yaml, pkg/providers/definitions/unstructured.yaml, pkg/providers/definitions/assemblyai.yaml - pkg/providers/schema.go - providers/huggingface.yaml (existing tier 3 reference — DO NOT MODIFY) Create each file in providers/ AND copy verbatim to pkg/providers/definitions/.

providers/perplexity.yaml:

format_version: 1
name: perplexity
display_name: Perplexity AI
tier: 3
last_verified: "2026-04-05"
keywords:
  - "perplexity"
  - "PERPLEXITY_API_KEY"
  - "pplx-"
  - "api.perplexity.ai"
patterns:
  - regex: 'pplx-[A-Za-z0-9]{48,}'
    entropy_min: 4.0
    confidence: high
verify:
  method: POST
  url: https://api.perplexity.ai/chat/completions
  headers:
    Authorization: "Bearer {KEY}"
  valid_status: [200, 400]
  invalid_status: [401, 403]

providers/you.yaml:

format_version: 1
name: you
display_name: You.com
tier: 3
last_verified: "2026-04-05"
keywords:
  - "you.com"
  - "YDC_API_KEY"
  - "YOU_API_KEY"
  - "api.ydc-index.io"
verify:
  method: GET
  url: https://api.ydc-index.io/search
  headers:
    X-API-Key: "{KEY}"
  valid_status: [200, 400]
  invalid_status: [401, 403]

providers/voyage.yaml:

format_version: 1
name: voyage
display_name: Voyage AI
tier: 3
last_verified: "2026-04-05"
keywords:
  - "voyage"
  - "VOYAGE_API_KEY"
  - "voyageai"
  - "api.voyageai.com"
patterns:
  - regex: 'pa-[A-Za-z0-9_\-]{40,}'
    entropy_min: 4.0
    confidence: medium
verify:
  method: POST
  url: https://api.voyageai.com/v1/embeddings
  headers:
    Authorization: "Bearer {KEY}"
  valid_status: [200, 400]
  invalid_status: [401, 403]

providers/jina.yaml:

format_version: 1
name: jina
display_name: Jina AI
tier: 3
last_verified: "2026-04-05"
keywords:
  - "jina"
  - "JINA_API_KEY"
  - "api.jina.ai"
  - "jinaai"
patterns:
  - regex: 'jina_[A-Za-z0-9]{40,}'
    entropy_min: 4.0
    confidence: high
verify:
  method: GET
  url: https://api.jina.ai/v1/models
  headers:
    Authorization: "Bearer {KEY}"
  valid_status: [200]
  invalid_status: [401, 403]

providers/unstructured.yaml:

format_version: 1
name: unstructured
display_name: Unstructured.io
tier: 3
last_verified: "2026-04-05"
keywords:
  - "unstructured"
  - "UNSTRUCTURED_API_KEY"
  - "api.unstructured.io"
  - "unstructuredio"
verify:
  method: GET
  url: https://api.unstructured.io/general/v0/general
  headers:
    unstructured-api-key: "{KEY}"
  valid_status: [200, 400]
  invalid_status: [401, 403]

providers/assemblyai.yaml:

format_version: 1
name: assemblyai
display_name: AssemblyAI
tier: 3
last_verified: "2026-04-05"
keywords:
  - "assemblyai"
  - "ASSEMBLYAI_API_KEY"
  - "api.assemblyai.com"
  - "assembly.ai"
patterns:
  - regex: '[a-f0-9]{32}'
    entropy_min: 4.0
    confidence: low
verify:
  method: GET
  url: https://api.assemblyai.com/v2/transcript
  headers:
    authorization: "{KEY}"
  valid_status: [200]
  invalid_status: [401, 403]

Copy all 6 files verbatim to pkg/providers/definitions/. cd /home/salva/Documents/apikey && for f in perplexity you voyage jina unstructured assemblyai; do diff providers/$f.yaml pkg/providers/definitions/$f.yaml || exit 1; done && go test ./pkg/providers/... -count=1 && go test ./pkg/engine/... -count=1 <acceptance_criteria> - All 12 files exist (6 providers × 2 locations) - grep -q 'pplx-' providers/perplexity.yaml - grep -q 'jina_' providers/jina.yaml - grep -q 'api.voyageai.com' providers/voyage.yaml - diff providers/perplexity.yaml pkg/providers/definitions/perplexity.yaml returns no diff - go test ./pkg/providers/... -count=1 passes - go test ./pkg/engine/... -count=1 passes </acceptance_criteria> 6 search/embeddings providers dual-located.

Task 2: Voice + image/video providers (Deepgram, ElevenLabs, Stability, Runway, Midjourney) providers/deepgram.yaml, providers/elevenlabs.yaml, providers/stability.yaml, providers/runway.yaml, providers/midjourney.yaml, pkg/providers/definitions/deepgram.yaml, pkg/providers/definitions/elevenlabs.yaml, pkg/providers/definitions/stability.yaml, pkg/providers/definitions/runway.yaml, pkg/providers/definitions/midjourney.yaml - pkg/providers/schema.go providers/deepgram.yaml: ```yaml format_version: 1 name: deepgram display_name: Deepgram tier: 3 last_verified: "2026-04-05" keywords: - "deepgram" - "DEEPGRAM_API_KEY" - "api.deepgram.com" - "dg-api" patterns: - regex: '[a-f0-9]{40}' entropy_min: 4.0 confidence: low verify: method: GET url: https://api.deepgram.com/v1/projects headers: Authorization: "Token {KEY}" valid_status: [200] invalid_status: [401, 403] ```

providers/elevenlabs.yaml:

format_version: 1
name: elevenlabs
display_name: ElevenLabs
tier: 3
last_verified: "2026-04-05"
keywords:
  - "elevenlabs"
  - "ELEVENLABS_API_KEY"
  - "ELEVEN_API_KEY"
  - "XI_API_KEY"
  - "api.elevenlabs.io"
patterns:
  - regex: '[a-f0-9]{32}'
    entropy_min: 4.0
    confidence: low
verify:
  method: GET
  url: https://api.elevenlabs.io/v1/user
  headers:
    xi-api-key: "{KEY}"
  valid_status: [200]
  invalid_status: [401, 403]

providers/stability.yaml:

format_version: 1
name: stability
display_name: Stability AI
tier: 3
last_verified: "2026-04-05"
keywords:
  - "stability"
  - "STABILITY_API_KEY"
  - "STABILITY_KEY"
  - "api.stability.ai"
  - "stable-diffusion"
patterns:
  - regex: 'sk-[A-Za-z0-9]{48}'
    entropy_min: 4.0
    confidence: medium
verify:
  method: GET
  url: https://api.stability.ai/v1/user/account
  headers:
    Authorization: "Bearer {KEY}"
  valid_status: [200]
  invalid_status: [401, 403]

providers/runway.yaml:

format_version: 1
name: runway
display_name: Runway
tier: 3
last_verified: "2026-04-05"
keywords:
  - "runway"
  - "runwayml"
  - "RUNWAY_API_KEY"
  - "RUNWAYML_API_SECRET"
  - "api.runwayml.com"
  - "dev.runwayml.com"
verify:
  method: GET
  url: https://api.dev.runwayml.com/v1/tasks
  headers:
    Authorization: "Bearer {KEY}"
  valid_status: [200, 400]
  invalid_status: [401, 403]

providers/midjourney.yaml:

format_version: 1
name: midjourney
display_name: Midjourney
tier: 3
last_verified: "2026-04-05"
keywords:
  - "midjourney"
  - "MIDJOURNEY_API_KEY"
  - "MJ_API_KEY"
  - "midjourney.com"
  - "useapi.net"
verify:
  method: GET
  url: ""
  headers: {}
  valid_status: []
  invalid_status: []

Copy all 5 files verbatim to pkg/providers/definitions/. cd /home/salva/Documents/apikey && for f in deepgram elevenlabs stability runway midjourney; do diff providers/$f.yaml pkg/providers/definitions/$f.yaml || exit 1; done && go test ./pkg/providers/... -count=1 && go test ./pkg/engine/... -count=1 <acceptance_criteria> - All 10 files exist - grep -q 'XI_API_KEY' providers/elevenlabs.yaml - grep -q 'stable-diffusion' providers/stability.yaml - grep -q 'runwayml' providers/runway.yaml - grep -L 'patterns:' providers/runway.yaml providers/midjourney.yaml returns both (keyword-only) - Total Tier 3 count: grep -l 'tier: 3' providers/*.yaml | wc -l → 12 (11 new + pre-existing huggingface) - go test ./pkg/providers/... -count=1 passes - go test ./pkg/engine/... -count=1 passes </acceptance_criteria> All 11 new Tier 3 Specialized providers dual-located. PROV-03 satisfied (12 total with pre-existing huggingface).

`grep -l 'tier: 3' providers/*.yaml | wc -l` returns 12.

<success_criteria>

• 11 new Tier 3 providers added (huggingface pre-existing, total 12)
• All dual-located
• Documented prefixes use tight regex; undocumented use keyword-only
• No engine regression </success_criteria>

After completion, create `.planning/phases/03-tier-3-9-providers/03-02-SUMMARY.md`