Update to Basic Memory Cloud

{
  "name": "memory-research",
  "description": "Research an external subject using web search, synthesize findings into a structured Basic Memory entity. Use when asked to research a company, person, technology, or topic — or when a bare name or URL is provided that implies a research request.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 361
    }
  ],
  "skill_md_contents": "---\nname: memory-research\ndescription: \"Research an external subject using web search, synthesize findings into a structured Basic Memory entity. Use when asked to research a company, person, technology, or topic — or when a bare name or URL is provided that implies a research request.\"\n---\n\n# Memory Research\n\nResearch an external subject, synthesize what you find, and create a structured Basic Memory entity — with the user's approval.\n\n## When to Use\n\n**Explicit triggers:**\n- \"Research [subject]\"\n- \"Look up [subject]\"\n- \"What do you know about [subject]?\"\n- \"Evaluate [subject]\"\n\n**Implicit triggers (also activate this skill):**\n- A bare name: \"Terraform\"\n- A URL: \"https://example.com\"\n- A name with context: \"Acme Corp — saw them at the conference\"\n\n## Workflow\n\n### Step 1: Web Research\n\nSearch for current information across multiple sources. Aim for 3-5 searches to build a well-rounded picture:\n\n```\n[subject name] site\n[subject name] overview\n[subject name] news [current year]\n[subject name] [relevant domain keywords]\n```\n\n**What to gather by entity type:**\n\n| Entity Type | Key Information |\n|-------------|----------------|\n| **Organization** | What they do, products/services, stage (startup/growth/public), funding, leadership, headquarters, employee count, notable partnerships or contracts |\n| **Person** | Current role, organization, background, expertise, notable work, public presence |\n| **Technology** | What it does, who maintains it, maturity, ecosystem, alternatives, adoption |\n| **Topic/Domain** | Definition, current state, key players, trends, relevance to user's context |\n\n### Step 2: Check Existing Knowledge\n\nBefore proposing a new entity, search Basic Memory:\n\n```python\nsearch_notes(query=\"Acme Corp\")\nsearch_notes(query=\"acme\")\n```\n\nTry name variations — full name, abbreviation, acronym, domain name.\n\nIf the entity already exists:\n- Report what you found in Basic Memory alongside your web research\n- Offer to update the existing note with new information\n- Use `edit_note` to append new observations or update outdated ones\n\nIf the entity doesn't exist, proceed to evaluation.\n\n### Step 3: Evaluate and Summarize\n\nPresent your findings in a structured summary. Include all relevant information organized by section:\n\n```markdown\n## [Subject Name]\n\n**Type:** [Organization / Person / Technology / Topic]\n\n**Summary:** [2-4 sentences: what this is, why it matters, key distinguishing facts]\n\n**Key Details:**\n- [Organized by what's relevant for the entity type]\n- [Stage, funding, leadership for orgs]\n- [Role, expertise, affiliations for people]\n- [Maturity, ecosystem, alternatives for tech]\n\n**Relevance:** [Why this matters to the user — connection to their work, domain, or interests.\nIf no obvious connection: \"No specific connection identified.\"]\n\n**Sources:**\n- [URLs of key sources consulted]\n```\n\n### Evaluation Guidelines\n\n**Use hedging language.** Web research is a snapshot, not ground truth:\n- \"Appears to be\", \"Based on public information\", \"Estimated\"\n- \"As of [date]\", \"According to [source]\"\n- Never state funding amounts, employee counts, or revenue as exact unless citing a primary source\n\n**Don't fabricate.** If information isn't available, say so:\n- \"Leadership information not publicly available\"\n- \"Funding details not disclosed\"\n\n**Let the user define relevance.** Don't impose a fixed evaluation framework. Instead, highlight facts and let the user draw conclusions. If the user has a specific evaluation rubric (strategic fit, buy/partner/compete, etc.), they'll tell you — apply it when asked.\n\n### Step 4: Propose Entity Creation\n\nAfter presenting the summary, ask for approval:\n\n```\nCreate Basic Memory entity for [Subject]?\n  Location: [suggested-folder]/[entity-name].md\n  Type: [entity type]\n\n  [yes / no / modify]\n```\n\nIf the user provided context with their request (\"saw them at the conference\"), include that context in the proposed entity.\n\n### Step 5: Create the Entity\n\nAfter approval, create a structured note. Adapt the template to the entity type:\n\n#### Organization\n\n```python\nwrite_note(\n  title=\"Acme Corp\",\n  directory=\"organizations\",\n  note_type=\"organization\",\n  tags=[\"organization\", \"relevant-tags\"],\n  content=\"\"\"# Acme Corp\n\n## Overview\n[2-3 sentence description from research]\n\n## Products & Services\n- [Key offerings discovered in research]\n\n## Background\n**Stage:** [Startup / Growth / Public]\n**Headquarters:** [Location]\n**Employees:** [Estimate, hedged]\n**Leadership:** [Key people if found]\n**Founded:** [Year if found]\n\n## Observations\n- [relevance] Why this entity matters in user's context\n- [source] Researched on YYYY-MM-DD\n- [additional observations from research findings]\n\n## Relations\n- [Link to related entities already in the knowledge graph]\"\"\"\n)\n```\n\n#### Person\n\n```python\nwrite_note(\n  title=\"Jane Smith\",\n  directory=\"people\",\n  note_type=\"person\",\n  tags=[\"person\", \"relevant-tags\"],\n  content=\"\"\"# Jane Smith\n\n## Overview\n[Current role and affiliation. Brief background.]\n\n## Background\n**Role:** [Title at Organization]\n**Expertise:** [Key domains]\n**Notable:** [Publications, talks, projects if found]\n\n## Observations\n- [role] Title at Organization\n- [expertise] Key technical or domain expertise\n- [source] Researched on YYYY-MM-DD\n\n## Relations\n- works_at [[Organization]]\"\"\"\n)\n```\n\n#### Technology\n\n```python\nwrite_note(\n  title=\"Technology Name\",\n  directory=\"concepts\",\n  note_type=\"concept\",\n  tags=[\"concept\", \"technology\", \"relevant-tags\"],\n  content=\"\"\"# Technology Name\n\n## Overview\n[What it is and what problem it solves]\n\n## Key Details\n**Maintained by:** [Organization or community]\n**Maturity:** [Experimental / Stable / Mature]\n**License:** [If applicable]\n**Alternatives:** [Comparable tools or approaches]\n\n## Observations\n- [definition] What this technology does in one sentence\n- [maturity] Current state and adoption level\n- [source] Researched on YYYY-MM-DD\n\n## Relations\n- [Link to related concepts, tools, or projects in the knowledge graph]\"\"\"\n)\n```\n\nAdapt these templates freely. The key elements are: note_type/tags parameters, an overview, structured details, observations with categories, and relations.\n\n### Step 6: Store Source Context\n\nIf the user provided context with their request, capture it in the entity:\n\n```python\n# User said: \"Acme Corp — saw their demo at the conference last week\"\nedit_note(\n  identifier=\"Acme Corp\",\n  operation=\"append\",\n  section=\"Observations\",\n  content=\"- [context] Saw their demo at conference, week of 2026-02-17\"\n)\n```\n\nThis context is often the most valuable part — it's the user's relationship to the entity, which web research can't provide.\n\n## Guidelines\n\n- **Always web search.** Don't rely on training data alone. Research should reflect current, verifiable information.\n- **Search Basic Memory first.** Check for existing entities before creating new ones. Update rather than duplicate.\n- **Hedge uncertain information.** Use qualifiers for estimates, unverified claims, and inferred details.\n- **Store source URLs.** Include the URLs you consulted, either in observations or a Sources section. This enables the user to verify and dig deeper.\n- **Get approval before creating.** Present your findings and let the user decide whether to create the entity and what to include.\n- **Capture user context.** If the user told you *why* they're researching (met at a conference, evaluating as a vendor, etc.), that context belongs in the entity.\n- **Don't over-research.** 3-5 web searches is usually enough. The goal is a useful knowledge graph entry, not an exhaustive report.\n- **Link to existing knowledge.** Relate the new entity to things already in the knowledge graph. Connections compound value.\n"
}