Skip to main content
Glama

recall_file

Retrieve the complete edit history of any file, with per-edit user-intent context. Understand why a file was modified historically by querying session_file_edits instead of summary memories.

Instructions

Get the COMPLETE edit history of a file across all sessions, with per-edit user-intent context. Returns: total edit count, daily breakdown, list of distinct user intents that drove the edits, and the linked memories. Use this when you need to understand WHY a file was modified historically — far more accurate than recall() for file-centric questions because it queries session_file_edits (every physical edit) instead of summary memories.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
path_substringYesSubstring to match against file_path (e.g. "search-services.ts" or full absolute path)
max_intentsNoMax distinct user-intent snippets to return. Default 10.

Implementation Reference

  • Tool registration for 'recall_file' in the TOOLS array, defining its name, description, and inputSchema (path_substring required, max_intents optional with default 10).
    {
      name: 'recall_file',
      description:
        'Get the COMPLETE edit history of a file across all sessions, with per-edit user-intent context. Returns: total edit count, daily breakdown, list of distinct user intents that drove the edits, and the linked memories. Use this when you need to understand WHY a file was modified historically — far more accurate than recall() for file-centric questions because it queries session_file_edits (every physical edit) instead of summary memories.',
      inputSchema: {
        type: 'object',
        properties: {
          path_substring: { type: 'string', description: 'Substring to match against file_path (e.g. "search-services.ts" or full absolute path)' },
          max_intents: { type: 'number', description: 'Max distinct user-intent snippets to return. Default 10.', default: 10 },
        },
        required: ['path_substring'],
      },
    },
  • Input schema for recall_file: path_substring (string, required) and max_intents (number, default 10).
    inputSchema: {
      type: 'object',
      properties: {
        path_substring: { type: 'string', description: 'Substring to match against file_path (e.g. "search-services.ts" or full absolute path)' },
        max_intents: { type: 'number', description: 'Max distinct user-intent snippets to return. Default 10.', default: 10 },
      },
      required: ['path_substring'],
    },
  • Main handler function handleRecallFile: queries session_file_edits table for total edit count, daily breakdown, user intents (context_snippets), linked memories, and matched file paths. Returns structured JSON with all results.
    function handleRecallFile(args: any): string {
      const sub = String(args.path_substring ?? '').trim();
      if (!sub) return JSON.stringify({ ok: false, error: 'path_substring required' });
      const maxIntents = Math.max(1, Math.min(50, args.max_intents ?? 10));
    
      const totalRow = db.prepare(`SELECT COUNT(*) as c, MIN(occurred_at) as first_at, MAX(occurred_at) as last_at, COUNT(DISTINCT session_id) as sessions FROM session_file_edits WHERE file_path LIKE ?`).get(`%${sub}%`) as any;
      if (!totalRow || totalRow.c === 0) {
        return JSON.stringify({ ok: true, count: 0, note: 'No edits found for that path substring.' });
      }
    
      // Daily breakdown
      const daily = db.prepare(`
        SELECT DATE(occurred_at, 'unixepoch') as day, operation, COUNT(*) as edits
        FROM session_file_edits WHERE file_path LIKE ?
        GROUP BY day, operation ORDER BY day
      `).all(`%${sub}%`) as any[];
    
      // Distinct context_snippets (intents) — deduped, ordered by recency
      const intents = db.prepare(`
        SELECT DISTINCT context_snippet, MAX(occurred_at) as last_at, COUNT(*) as freq
        FROM session_file_edits
        WHERE file_path LIKE ? AND context_snippet IS NOT NULL AND LENGTH(context_snippet) > 20
        GROUP BY context_snippet
        ORDER BY last_at DESC
        LIMIT ?
      `).all(`%${sub}%`, maxIntents) as any[];
    
      // Linked memories
      const memories = db.prepare(`
        SELECT DISTINCT m.id, m.layer, m.content, m.importance, e.name as entity_name
        FROM session_file_edits sfe
        JOIN memories m ON m.id = sfe.memory_id
        JOIN entities e ON e.id = m.entity_id
        WHERE sfe.file_path LIKE ?
        ORDER BY m.importance DESC
        LIMIT 20
      `).all(`%${sub}%`) as any[];
    
      // Distinct file paths matched (the substring may match multiple files)
      const paths = db.prepare(`SELECT file_path, COUNT(*) as edits FROM session_file_edits WHERE file_path LIKE ? GROUP BY file_path ORDER BY edits DESC`).all(`%${sub}%`) as any[];
    
      return JSON.stringify({
        ok: true,
        path_substring: sub,
        paths_matched: paths,
        summary: {
          total_edits: totalRow.c,
          first_edit_at: new Date(totalRow.first_at * 1000).toISOString(),
          last_edit_at: new Date(totalRow.last_at * 1000).toISOString(),
          sessions_involved: totalRow.sessions,
        },
        daily_breakdown: daily,
        user_intents: intents.map((i) => ({
          when: new Date(i.last_at * 1000).toISOString(),
          occurrences: i.freq,
          intent: i.context_snippet,
        })),
        linked_memories: memories.map((m) => ({
          id: m.id,
          entity: m.entity_name,
          layer: m.layer,
          importance: m.importance,
          preview: m.content.length > 300 ? m.content.slice(0, 300) + '...' : m.content,
        })),
      });
    }
  • MCP CallToolRequestSchema dispatcher: routes 'recall_file' to handleRecallFile function.
    case 'recall_file': text = handleRecallFile(args); break;
  • Telemetry tracking: recall_file_calls field in TelemetryPayload type, initialized to 0 in buildPayload.
    recall_file_calls: number;
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations provided, so description carries full burden. It discloses what data is returned (edit count, daily breakdown, intents, linked memories) and that it queries session_file_edits. However, it does not address potential issues like multiple file matches or performance implications.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Two sentences plus return list: efficient, front-loaded with key information, no wasted words.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given 2 params, no output schema, and no annotations, the description explains return values (edit count, daily breakdown, intents, linked memories) but lacks details on format, error handling, or behavior for multiple matches.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% with both parameters already described. The description does not add significant new meaning beyond what the schema provides, so baseline 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Description clearly states verb 'Get' and resource 'edit history of a file', emphasizes 'COMPLETE' and 'per-edit user-intent context', and distinguishes from sibling recall() by specifying it queries session_file_edits.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Explicitly tells when to use: 'when you need to understand WHY a file was modified historically' and contrasts with recall(), providing clear guidance on tool selection.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/michielinksee/linksee-memory'

If you have feedback or need assistance with the MCP directory API, please join our Discord server