memory-protocol
Universal protocol for total-agent-memory MCP server. Activate at session start, before any non-trivial task, after every…
Activity report from total-agent-memory for a project and period. Use when the user types /report or asks "report for today", "what did we do this week", "weekly / monthly summary", "status report", "retro for project X", "all-time report", "report for last week", or a report
Claude Code
$ npx -y skills add vbcherepanov/total-agent-memory --skill report --agent claude-code
Codex
$ npx skills add vbcherepanov/total-agent-memory --skill report -a codex -y
Cursor
$ npx skills add vbcherepanov/total-agent-memory --skill report -a cursor -y
Gemini CLI
$ npx skills add vbcherepanov/total-agent-memory --skill report -a gemini-cli -y
GitHub Copilot
$ npx skills add vbcherepanov/total-agent-memory --skill report -a github-copilot -y
OpenCode
$ npx skills add vbcherepanov/total-agent-memory --skill report -a opencode -y
Windsurf
$ npx skills add vbcherepanov/total-agent-memory --skill report -a windsurf -y
name: report description: > Activity report from total-agent-memory for a project and period. Use when the user types /report or asks "report for today", "what did we do this week", "weekly / monthly summary", "status report", "retro for project X", "all-time report", "report for last week", or a report for a department or the whole company on the team server. Calls the memory_report MCP tool and presents the result. argument-hint: "[today|week|month|all|YYYY-MM-DD..YYYY-MM-DD] [project]"
The report is built by the `memory_report` tool (your client may prefix it, e.g. `mcp__memory__memory_report`). It is deterministic: numbers, decisions, fixes, errors, open next steps, files and a daily timeline come straight from stored records, each with source IDs. If no `memory_report` tool is available, say the memory server is not connected (or is older than this skill) and stop.
Take the period and project from `$ARGUMENTS` or the user's words:
| The user says | Arguments | |---|---| | today, for the day, daily | `period: "day"` | | yesterday | `period: "day", offset: -1` | | this week, weekly (default when nothing is said) | `period: "week"` | | last week | `period: "week", offset: -1` | | this month, monthly | `period: "month"` | | last month | `period: "month", offset: -1` | | all time, ever, since the start | `period: "all"` | | from 2026-09-01 to 2026-09-15, since Monday, last 10 days | `period: "custom", since: "YYYY-MM-DD", until: "YYYY-MM-DD"` (until is inclusive; omit it for "until today") |
user says "this project". Omit `project` for "all projects".
zone or city; otherwise the server uses the machine's zone.
"company report" → `scope: "company"`; otherwise `scope: "personal"`. A `Forbidden` error means the user lacks the department head / company viewer role: say so plainly.
Call `memory_report` with those arguments and `format: "markdown"`. Add `include_llm_summary: true` only when the user asks for a written summary or narrative; it needs a configured LLM and the report is complete without it.
Do not paste the whole report unless asked. Give:
1. One line with the period (as the report labels it) and the headline numbers with their change against the previous period (records, decisions, errors, active days). 2. The 3–5 most important decisions with their WHY, and the main fixes. 3. Recurring error patterns (count this period vs all time) and lessons learned. 4. Open next steps and pitfalls that were not picked up yet. 5. Anything notable in files, technologies or the timeline (e.g. one very busy day).
Keep the `#id` references so the user can ask for details; fetch full records with `memory_get(ids=[...])` when they do (on the team server: `memory_get` with the scope and `id`). If the report says there was no activity, say so and offer a longer period.
End by offering to save it. On yes, call `memory_report` again with the same arguments and `save: true`, then show the returned file path (`<memory dir>/reports/<project>/<period>-<date>.md`). The same report is available from a terminal: `tam report --project X --period week [--format json] [--out FILE]`.
Persistent local memory for AI coding agents — Claude Code, Codex CLI, Cursor, any MCP client. Temporal knowledge graph, procedural memory, AST codebase ingest, cross-project analogy. LongMemEval R@5 95.1%, LoCoMo 0.607, BEAM 1M 0.448. 74 MCP tools, 9 IDEs, 100% local.
Universal protocol for total-agent-memory MCP server. Activate at session start, before any non-trivial task, after every…
Department head / company viewer view of employee onboarding on the total-agent-memory team server. Use when the user types…
Department onboarding for a new employee on the total-agent-memory team server. Use when the user types /onboard, asks to be…
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or…
Ask which skill or flow fits your situation. A router over the skills in this repo.