# Gather Context > Gather relevant context for a task using code search, knowledge base, and known issues. Use at the start of skills that need context about a Jira issue, feature, bug, or code review. Use when this capability is needed. - Skill: `tomevault-io/gather-context` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/gather-context` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/gather-context/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Research & Search - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/gather-context --- # Gather Context Consolidates: code_search, knowledge_query, and check_known_issues into structured context. ## Inputs | Input | Type | Default | Purpose | |-------|------|---------|---------| | `query` | string | required | What to search (issue description, feature name, error message) | | `project` | string | "automation-analytics-backend" | Project for knowledge lookup | | `tool_name` | string | - | Tool to check for known issues | | `code_limit` | int | 5 | Max code search results | | `include_architecture` | bool | true | Include architecture overview | | `include_gotchas` | bool | true | Include project gotchas | | `include_patterns` | bool | true | Include coding patterns | ## Workflow ### 1. Code Search - `code_search(query=query, project=project, limit=code_limit)` - Parse into structured results (file paths, relevance) ### 2. Load Gotchas (if include_gotchas) - `knowledge_query(project=project, section="gotchas")` - Parse list items into actionable gotchas ### 3. Load Patterns (if include_patterns) - `knowledge_query(project=project, section="patterns.coding")` - Parse into coding patterns list ### 4. Load Architecture (if include_architecture) - `knowledge_query(project=project, section="architecture.overview")` ### 5. Check Known Issues - `check_known_issues(tool_name=tool_name, error_text=query[:100])` - Parse into known issues list ### 6. Build Combined Context Output object with: - `code`: found, count, results - `gotchas`: found, count, items - `patterns`: found, count, items - `architecture`: found, overview - `known_issues`: found, count, items - `summary`: has_code, has_gotchas, has_patterns, total_items ### 7. Log - `memory_session_log("Gathered context for: {query[:50]}", "Code: X, Gotchas: Y, Patterns: Z")` ## Key MCP Tools - `code_search` — semantic code search - `knowledge_query` — gotchas, patterns, architecture - `check_known_issues` — known fixes for tools/errors - `memory_session_log` — session logging --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/dmzoneill) — claim your Tome and manage your conversions.