hive

List vault projects, or files within a project.

When called without arguments, lists all available projects. When called with a project, lists files in that project directory.

Args: project: Project slug. Empty = list all projects. path: Subdirectory within the project. Empty = project root. pattern: Glob pattern to filter files (e.g. 'adr-', '.md').

Read content from a vault project — use instead of direct filesystem access.

Args: project: Project slug (directory under 10_projects/), or '_meta' for 00_meta/. section: Shortcut name (context, tasks, roadmap, lessons). Ignored if path is set. path: Relative path to a specific .md file within the project. Overrides section. max_lines: Maximum lines to return. 0 = unlimited. include_metadata: Prepend a structured metadata line from YAML frontmatter.

Search the vault: full-text, ranked, or recent changes.

Default mode: flat full-text search across all vault files. Ranked mode (ranked=True): results scored by relevance. Recent mode (since_days>0): files changed in the last N days.

Args: query: Text to search for (case-insensitive). max_lines: Maximum output lines. Default 500. type_filter: Only files whose frontmatter type matches. status_filter: Only files whose frontmatter status matches. tag_filter: Only files that have this frontmatter tag. use_regex: Treat query as regex. Default False. ranked: Score results by relevance. Default False. max_results: Max files when ranked. Default 10. since_days: Show recent changes (0 = disabled). Default 0. project: Filter to this project (recent mode only).

Call at the start of every new session to load project context.

Assembles active tasks, recent lessons, git activity, and project health into a single response — replaces 3-4 manual tool calls.

Args: project: Project slug (directory under 10_projects/).

Write to the vault: append, replace a section, or create a new file.

Modes:

• append/replace: Update a project section. Requires section.

• create: Create a new file with auto-generated frontmatter. Requires path and doc_type.

Args: project: Project slug or '_meta' for cross-project content. content: Markdown content to write (body only for create mode). operation: 'append', 'replace', or 'create'. Default 'append'. section: Section shortcut (context, tasks, roadmap, lessons). For append/replace. path: Relative path for new file. For create mode. doc_type: Document type for frontmatter. For create mode.

Surgical text replacement in a vault file with auto git commit.

Supports single or multi-replacement. For single replacement, provide old_text and new_text. For multiple replacements, provide patches — a list of {old_text, new_text} dicts applied in sequence. Do not mix both modes.

Each old_text must appear exactly once in the file (after prior patches in the list have been applied). If any patch fails validation, no changes are written.

Args: project: Project slug or '_meta' for cross-project content. path: Relative path to the file within the project. old_text: Exact text to find and replace (single mode). Empty = not set. new_text: Replacement text (single mode). Empty = not set. patches: List of {old_text, new_text} dicts (multi mode).

Return vault health metrics, validation, and optional usage analytics.

Without parameters, returns a health summary for all projects. When checks are specified, runs drift detection (frontmatter, stale, links). When include_usage is True, appends tool usage analytics.

Args: project: Project slug to validate. Empty = all projects. checks: Validation checks to run. Empty = health summary only. Options: frontmatter, stale, links. max_issues: Maximum validation issues to report. Default 50. include_usage: Append vault tool usage analytics. Default False. usage_days: Usage look-back window in days. Default 30.

Capture lessons: inline (structured fields) or batch (raw text via worker).

Inline mode (default): provide title, context, problem, solution. Batch mode: provide text to extract lessons automatically via worker.

Args: project: Project slug (directory under 10_projects/). title: Short descriptive title (inline mode). context: What you were doing (inline mode). problem: What went wrong or what decision was needed (inline mode). solution: What fixed it or what was decided (inline mode). tags: Optional tags (e.g. ["python", "testing"]). text: Raw text to extract lessons from (batch mode). min_confidence: Minimum confidence for batch extraction. Default 0.7. max_lessons: Maximum lessons to extract in batch mode. Default 5.

Offload work to a cheaper model or summarize vault files.

When project is provided, reads a vault file. Small files (≤50 lines) are returned directly. Large files are auto-delegated to a worker for summarization — falls back to raw content if workers are unavailable.

Args: prompt: The task description or code to process. context: Optional system context for the model. model: 'auto', 'ollama', 'openrouter-free', 'openrouter' (paid), or model ID. max_tokens: Maximum tokens in the response. max_cost_per_request: Max USD. 0 = free models only. project: Project slug for vault summarization mode. section: Shortcut name for summarization. Ignored if path is set. path: Relative path to a .md file. Overrides section. max_summary_lines: Target summary length for summarization.

Show worker health: budget, connectivity, available models, and usage stats.

Args: include_models: Include available model list from all providers. Default True.