AgentRecall-MCP

Name: AgentRecall-MCP Review
Item: AgentRecall-MCP
Rating: 3.2
Author: Aaron Matthews

Connector Development Usable

Works inClaude Code Claude Cowork Claude Chat

Usable Scanned — metadata only

Agents retain context and learnings across sessions, reducing redundant analysis and improving decision quality on iterative tasks.

An MCP server that gives Claude Code persistent memory built around a corrections ledger: each time you correct the agent, the correction is stored with severity, evidence, and outcome tracking, and later sessions record whether it was heeded or recurred. Ships as an MCP server, SDK, and CLI (npm: agent-recall-mcp), local-first with no cloud by default. Unusually, the project publishes its own measured recall and heed-rate numbers, including unflattering ones.

Code review agent learning from previous PR feedback patterns
Research agent building knowledge across multiple document analysis sessions
Debugging assistant refining hypotheses from failed test runs

307 starsMIT (commercial OK)FreeNo code needed

Usable rating — This tool is functional but has notable gaps. Review the evaluation notes below before deploying.

Why it matters

Agents retain context and learnings across sessions, reducing redundant analysis and improving decision quality on iterative tasks. Think-Execute-Reflect loops let Claude refine its approach based on previous outcomes.

Best for

Teams running multi-turn agent workflows where consistency and accumulated context reduce rework—code review agents, research synthesis, or debugging loops.

Works in

Claude Code Claude Cowork Claude Chat

Source

https://github.com/Goldentrii/AgentRecall-MCP-MCP

By Goldentrii

How to Get It

Option 1: Claude Desktop AppOpen the Customize panel in the sidebar → browse connectors → search and add. Works in Claude Code, Claude Cowork, and Claude Chat.

Option 2: Paste into Claude CodeCopy the command below and paste it into a Claude Code conversation. Claude will run it for you.

Command

claude mcp add --scope user agent-recall -- npx -y agent-recall-mcp

Tip: Paste this into a Claude Code conversation. Verify command matches your Claude Code version.

First thing to try

Once it’s connected, paste this into Claude:

Help me code review agent learning from previous PR feedback patterns

CostFree

Trust Signals Auto-scanned

Stars307Contributors1Last updated2026-07-05LicenseMIT (OK for commercial use)Known CVEsNone foundSources: GitHub Advisory Database + OSV.dev · Scanned 2026-08-13 · scanner vattempted-no-data

Data & Access

Data processingPrompts sent to Anthropic API. Enterprise/Team plans exclude training.

Reviewer notes

Auto-scanned review. These are observations, not a security certification.

catalog_hygiene stale-eval refresh: Scored from trust signals (evidence-eval-v1): 307 GitHub stars; 1 contributors; last commit 44d ago; license MIT.

Things to check

Scanned, not hands-on tested — this entry was auto-scanned from public metadata (GitHub metrics, license, security flags). No reviewer has run it, and no tool-specific limitations have been documented yet.
Single maintainer. Consider the risk if this person stops maintaining the project.

How to evaluate tools before deploying →

Data shown here comes from public APIs and automated scanning. Reviewer notes reflect one person's experience. This is not a security certification or legal recommendation. Always evaluate tools according to your own organization's policies.

Evaluation Scores

Evaluation

Ease of Use

3/5

Versatility

4/5

Reliability

3/5

Security

3/5

Overall score3.25 / 5.00 UsableEvaluatedAug 2026

catalog_hygiene stale-eval refresh: Scored from trust signals (evidence-eval-v1): 307 GitHub stars; 1 contributors; last commit 44d ago; license MIT.

← Back to the Claude Observatory