feat: updated agent harness

This commit is contained in:
DESKTOP-RTLN3BA\$punk 2026-04-28 09:22:19 -07:00
parent 9ec9b64348
commit 31a372bb84
139 changed files with 12583 additions and 1111 deletions

View file

@ -0,0 +1,12 @@
- User: "Generate an image of a cat"
- Call: `generate_image(prompt="A fluffy orange tabby cat sitting on a windowsill, bathed in warm golden sunlight, soft bokeh background with green houseplants, photorealistic style, cozy atmosphere")`
- The generated image will automatically be displayed in the chat.
- User: "Draw me a logo for a coffee shop called Bean Dream"
- Call: `generate_image(prompt="Minimalist modern logo design for a coffee shop called 'Bean Dream', featuring a stylized coffee bean with dream-like swirls of steam, clean vector style, warm brown and cream color palette, white background, professional branding")`
- The generated image will automatically be displayed in the chat.
- User: "Show me this image: https://example.com/image.png"
- Simply include it in your response using markdown: `![Image](https://example.com/image.png)`
- User uploads an image file and asks: "What is this image about?"
- The user's uploaded image is already visible in the chat.
- Simply analyze the image content and respond directly.

View file

@ -0,0 +1,7 @@
- User: "Give me a podcast about AI trends based on what we discussed"
- First search for relevant content, then call: `generate_podcast(source_content="Based on our conversation and search results: [detailed summary of chat + search findings]", podcast_title="AI Trends Podcast")`
- User: "Create a podcast summary of this conversation"
- Call: `generate_podcast(source_content="Complete conversation summary:\n\nUser asked about [topic 1]:\n[Your detailed response]\n\nUser then asked about [topic 2]:\n[Your detailed response]\n\n[Continue for all exchanges in the conversation]", podcast_title="Conversation Summary")`
- User: "Make a podcast about quantum computing"
- First explore `/documents/` (ls/glob/grep/read_file), then: `generate_podcast(source_content="Key insights about quantum computing from retrieved files:\n\n[Comprehensive summary of findings]", podcast_title="Quantum Computing Explained")`

View file

@ -0,0 +1,13 @@
- User: "Generate a report about AI trends"
- Call: `generate_report(topic="AI Trends Report", source_strategy="kb_search", search_queries=["AI trends recent developments", "artificial intelligence industry trends", "AI market growth and predictions"], report_style="detailed")`
- WHY: Has creation verb "generate" → call the tool. No prior discussion → use kb_search.
- User: "Write a research report from this conversation"
- Call: `generate_report(topic="Research Report", source_strategy="conversation", source_content="Complete conversation summary:\n\n...", report_style="deep_research")`
- WHY: Has creation verb "write" → call the tool. Conversation has the content → use source_strategy="conversation".
- User: (after a report on Climate Change was generated) "Add a section about carbon capture technologies"
- Call: `generate_report(topic="Climate Crisis: Causes, Impacts, and Solutions", source_strategy="conversation", source_content="[summary of conversation context if any]", parent_report_id=<previous_report_id>, user_instructions="Add a new section about carbon capture technologies")`
- WHY: Has modification verb "add" + specific deliverable target → call the tool with parent_report_id.
- User: (after a report was generated) "What else could we add to have more depth?"
- Do NOT call generate_report. Answer in chat with suggestions.
- WHY: No creation/modification verb directed at producing a deliverable.

View file

@ -0,0 +1,19 @@
- User: "Build me a resume. I'm John Doe, engineer at Acme Corp..."
- Call: `generate_resume(user_info="John Doe, engineer at Acme Corp...", max_pages=1)`
- WHY: Has creation verb "build" + resume → call the tool.
- User: "Create my CV with this info: [experience, education, skills]"
- Call: `generate_resume(user_info="[experience, education, skills]", max_pages=1)`
- User: "Build me a resume" (and there is a resume/CV document in the conversation context)
- Extract the FULL content from the document in context, then call:
`generate_resume(user_info="Name: John Doe\nEmail: john@example.com\n\nExperience:\n- Senior Engineer at Acme Corp (2020-2024)\n Led team of 5...\n\nEducation:\n- BS Computer Science, MIT (2016-2020)\n\nSkills: Python, TypeScript, AWS...", max_pages=1)`
- WHY: Document content is available in context — extract ALL of it into user_info. Do NOT ignore referenced documents.
- User: (after resume generated) "Change my title to Senior Engineer"
- Call: `generate_resume(user_info="", user_instructions="Change the job title to Senior Engineer", parent_report_id=<previous_report_id>, max_pages=1)`
- WHY: Modification verb "change" + refers to existing resume → set parent_report_id.
- User: (after resume generated) "Make this 2 pages and expand projects"
- Call: `generate_resume(user_info="", user_instructions="Expand projects and keep this to at most 2 pages", parent_report_id=<previous_report_id>, max_pages=2)`
- WHY: Explicit page increase request → set max_pages to 2.
- User: "How should I structure my resume?"
- Do NOT call generate_resume. Answer in chat with advice.
- WHY: No creation/modification verb.

View file

@ -0,0 +1,7 @@
- User: "Give me a presentation about AI trends based on what we discussed"
- First search for relevant content, then call: `generate_video_presentation(source_content="Based on our conversation and search results: [detailed summary of chat + search findings]", video_title="AI Trends Presentation")`
- User: "Create slides summarizing this conversation"
- Call: `generate_video_presentation(source_content="Complete conversation summary:\n\nUser asked about [topic 1]:\n[Your detailed response]\n\nUser then asked about [topic 2]:\n[Your detailed response]\n\n[Continue for all exchanges in the conversation]", video_title="Conversation Summary")`
- User: "Make a video presentation about quantum computing"
- First explore `/documents/` (ls/glob/grep/read_file), then: `generate_video_presentation(source_content="Key insights about quantum computing from retrieved files:\n\n[Comprehensive summary of findings]", video_title="Quantum Computing Explained")`

View file

@ -0,0 +1,13 @@
- User: "Check out https://dev.to/some-article"
- Call: `scrape_webpage(url="https://dev.to/some-article")`
- Respond with a structured analysis — key points, takeaways.
- User: "Read this article and summarize it for me: https://example.com/blog/ai-trends"
- Call: `scrape_webpage(url="https://example.com/blog/ai-trends")`
- Respond with a thorough summary using headings and bullet points.
- User: (after discussing https://example.com/stats) "Can you get the live data from that page?"
- Call: `scrape_webpage(url="https://example.com/stats")`
- IMPORTANT: Always attempt scraping first. Never refuse before trying the tool.
- User: "https://example.com/blog/weekend-recipes"
- Call: `scrape_webpage(url="https://example.com/blog/weekend-recipes")`
- When a user sends just a URL with no instructions, scrape it and provide a concise summary of the content.

View file

@ -0,0 +1,9 @@
- User: "How do I install SurfSense?"
- Call: `search_surfsense_docs(query="installation setup")`
- User: "What connectors does SurfSense support?"
- Call: `search_surfsense_docs(query="available connectors integrations")`
- User: "How do I set up the Notion connector?"
- Call: `search_surfsense_docs(query="Notion connector setup configuration")`
- User: "How do I use Docker to run SurfSense?"
- Call: `search_surfsense_docs(query="Docker installation setup")`

View file

@ -0,0 +1,16 @@
- <user_name>Alex</user_name>, <user_memory> is empty. User: "I'm a space enthusiast, explain astrophage to me"
- The user casually shared a durable fact. Use their first name in the entry, short neutral heading:
update_memory(updated_memory="## Interests & background\n- (2025-03-15) [fact] Alex is a space enthusiast\n")
- User: "Remember that I prefer concise answers over detailed explanations"
- Durable preference. Merge with existing memory, add a new heading:
update_memory(updated_memory="## Interests & background\n- (2025-03-15) [fact] Alex is a space enthusiast\n\n## Response style\n- (2025-03-15) [pref] Alex prefers concise answers over detailed explanations\n")
- User: "I actually moved to Tokyo last month"
- Updated fact, date prefix reflects when recorded:
update_memory(updated_memory="## Interests & background\n...\n\n## Personal context\n- (2025-03-15) [fact] Alex lives in Tokyo (previously London)\n...")
- User: "I'm a freelance photographer working on a nature documentary"
- Durable background info under a fitting heading:
update_memory(updated_memory="...\n\n## Current focus\n- (2025-03-15) [fact] Alex is a freelance photographer\n- (2025-03-15) [fact] Alex is working on a nature documentary\n")
- User: "Always respond in bullet points"
- Standing instruction:
update_memory(updated_memory="...\n\n## Response style\n- (2025-03-15) [instr] Always respond to Alex in bullet points\n")

View file

@ -0,0 +1,7 @@
- User: "Let's remember that we decided to do weekly standup meetings on Mondays"
- Durable team decision:
update_memory(updated_memory="- (2025-03-15) [fact] Weekly standup meetings on Mondays\n...")
- User: "Our office is in downtown Seattle, 5th floor"
- Durable team fact:
update_memory(updated_memory="- (2025-03-15) [fact] Office location: downtown Seattle, 5th floor\n...")

View file

@ -0,0 +1,8 @@
- User: "What's the current USD to INR exchange rate?"
- Call: `web_search(query="current USD to INR exchange rate")`
- Then answer using the returned web results with citations.
- User: "What's the latest news about AI?"
- Call: `web_search(query="latest AI news today")`
- User: "What's the weather in New York?"
- Call: `web_search(query="weather New York today")`