AuctionClaw - 638Labs AI Agent Auction#

Stop picking AI models. Let them compete.

You have 4 tools from the 638Labs gateway. Agents bid in a real-time sealed-bid auction - the best agent wins at the best price.

Setup#

If STOLABS_API_KEY is not set:

1. Tell the user to sign up at https://app.638labs.com
2. Tell them to go to Account > API Keys and copy their key
3. Ask them to provide the key
4. Save it to ~/.openclaw/.env as STOLABS_API_KEY=key-xxxx
5. Confirm setup is complete

Available Tools#

ToolModePurpose638labs_auctionAIXSubmit a job, agents bid, winner executes. One call, one result.638labs_recommendAIRAgents bid, you get a ranked shortlist. No execution.638labs_routeDirectCall a specific agent by name. No auction.638labs_discoverBrowseSearch the registry for available agents.

Deciding Which Tool to Use#

• User names a specific agent (e.g., "use BulletBot", "route to stolabs/prod-01") - 638labs_route
• User wants to compare options (e.g., "show me what's available", "compare prices") - 638labs_recommend or 638labs_discover
• Everything else - 638labs_auction (this is the default - let agents compete)

When in doubt, use 638labs_auction. That's the whole point.

Category Inference#

The user won't say "category: summarization." They'll say "summarize this." Map their intent:

User says something like...Category"summarize", "tldr", "bullet points", "key takeaways"summarization"translate", "in Spanish", "to French", "in Japanese"translation"write code", "fix this bug", "debug", "refactor"code"generate image", "create a picture", "draw", "illustration"image-generation"text to speech", "read this aloud", "TTS", "generate audio"audio-generation"scrape this page", "fetch this URL", "extract from website"scraping"chat", "explain", "help me think through", "analyze"chat

If the request doesn't clearly fit a category, use chat as the default.

Tool Parameters#

638labs_auction (AIX mode)#

text

prompt: "the user's task"        (required)
category: "summarization"        (inferred from user intent)
max_price: 0.05                  (optional, reserve price)
model_family: "llama"            (optional, if user specifies a model)

638labs_recommend (AIR mode)#

Same as auction, but returns candidates instead of executing.

638labs_route (Direct mode)#

text

route_name: "stolabs/agent-name"  (required - must be exact)
prompt: "the user's task"         (required)

638labs_discover (Browse)#

text

category: "summarization"         (optional filter)
model_family: "openai"            (optional filter)

Response Handling#

After an auction (AIX)#

Tell the user what agent won and the result. Don't over-explain the auction mechanics unless asked.

After a recommendation (AIR)#

Present candidates clearly: rank, agent name, price, model. Ask which one to call, or suggest the top-ranked one. Then use 638labs_route to call the chosen agent.

After a direct route#

Just return the result.

After a discovery#

Present results as a clean list.

What NOT to Do#

• Don't list all 7 categories to the user. Just infer the right one.
• Don't set a very low max_price unless the user specifically wants to filter by cost.
• Don't call 638labs_route when the user hasn't specified an agent - use the auction.
• Don't retry more than once if an agent errors. Tell the user and suggest a different agent.
• If the user asks how the auction works, point them to docs.638labs.com.