mcp-memory-service is a self-hosted memory backend that lets AI agents store and retrieve shared project context through REST, MCP, OAuth, a command-line interface, and a dashboard. It is intended for agent pipelines and clients such as LangGraph, CrewAI, AutoGen, Claude Desktop, and OpenCode, with support for knowledge graphs and memory consolidation. The catalogue includes skills, agents, commands, instructions, hooks, a setting, an MCP entry, and a plugin for its workflows.
Borrowing it
Nothing to install: this file belongs to doobidoo/mcp-memory-service. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/doobidoo/mcp-memory-service/main/.claude/commands/merge-sweep.mdgit clone --depth 1 https://github.com/doobidoo/mcp-memory-serviceWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/doobidoo/mcp-memory-service/merge-sweep)<a href="https://agentmods.dev/commands/doobidoo/mcp-memory-service/merge-sweep"><img src="https://agentmods.dev/badge/commands/doobidoo/mcp-memory-service/merge-sweep/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/doobidoo/mcp-memory-service/merge-sweep"><img src="https://agentmods.dev/badge/commands/doobidoo/mcp-memory-service/merge-sweep.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.00511 |
| Opus 5.5 | $0.00000 | $0.00204 |
| Sonnet 5.5 | $0.00000 | $0.00102 |
| Haiku 4.5 | $0.00000 | $0.00051 |
Grade A, and why
merge-sweep scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 15d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
/merge-sweep — Merge Open PRs With Minimal Effort
Role: merge coordinator for open PRs (GitHub, gh). Goal: merge as many PRs as
possible with minimal effort.
1. Inventory
gh pr list --state open --json number,title,author,headRefOid,isDraft,mergeable,mergeStateStatus,labels
2. Check each PR — read only, change nothing
- CI per head SHA (
gh api repos/{owner}/{repo}/commits/<sha>/check-runs), notgh pr checks. - Unresolved review threads (GraphQL
reviewThreads,isResolved). ResolveThreads has no bypass. - Conflicts or drift against main. A fork PR showing
action_requiredmeans the run needs approval. - Dependabot: did the bump widen a load-bearing constraint?
Only when a question needs a local answer: worktree under ../worktrees/pr-<N>, run
.venv/bin/pytest scoped there. Never touch the main tree — several sessions share it.
3. Classify
- MERGE: green, no open threads, no conflicts, automated-review findings (Greptile
summary and confidence) read, and for contributor PRs the diff reviewed (
/pr-review) - QUICK-FIX: under 10 minutes (rebase, answer a thread, approve a run) — name the fix
- DECISION: needs the maintainer's judgment
- SKIP: draft, red for a real reason, large
4. Report
One table: PR | title | category | reason (one line) | evidence command. Then STOP. Merge nothing.
5. After approval
- Merge one at a time:
gh pr merge <N> --squash --admin --match-head-commit <sha>. Immediately before each merge, re-fetch the head SHA, its threads and its check-runs; if the SHA changed or a required check is not green, stop and re-classify that PR. - Strict policy: after each merge, update the next PR and wait for green CI.
- Do not run
collect_changelog.py— collecting fragments belongs to/release. - Finally remove the worktrees and sync the GitLab mirror with a proven fast-forward (see CLAUDE.md "Source Control & Hosting").
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 15d ago First seen · 44 lines · 0 tokens per session scan A 9b8f77e33258
merge-sweep is a command published in the GitHub repository doobidoo/mcp-memory-service (2,021 stars, last pushed yesterday), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 511 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-04.
Other commands, from other repositories
context
A project-context command that manages a .context directory containing decision notes, session logs, and history. It helps keep a record of why changes were made across Git branches and commits.
bm
Ask basemind anything about the current codebase — outlines, refs, callers, git history, blame, diffs, docs, memory.
context-load
Carga de contexto al inicio de sesión. Lee estado del workspace, decisiones recientes, último session save y actividad Git para arrancar con el big picture.
catchup
A command that rebuilds an agent's working context from project notes and recent Git activity.
review-code
Read ALL memory bank code rules + best practices, check the files that changed, and APPLY fixes so they adhere. The active counterpart to /scan (which is read-only). Use after an AI session, before commit, to make changed files compliant.
yolo-abort
Immediate halt of /yolo Hive session mid-action — partial state saved, in-flight git operations rolled back where possible, audit log finalized with abort flag.