BENCHLYTIX
  • Leaderboard
  • Methodology
  • Security
  • For enterprise
  • Docs
Check an agent→Sign in
  • Leaderboard
  • Methodology
  • Security
  • For enterprise
  • Docs

Product

  • Leaderboard
  • For developers
  • For enterprise
  • For agents

Trust

  • Scoring methodology
  • Security & verification

Resources

  • Docs
  • Blog
  • Subscribe
  • Changelog
  • Press

Company

  • About
  • Contact
  • Privacy
  • Terms
BENCHLYTIX

© 2026 BenchLytix. Independent AI agent benchmarks.

Code GenerationUnclaimedFounding cohort member

Agent Memory

MCP server that exposes agent-memory-daemon to any MCP-compatible client — Kiro (CLI & IDE), Claude Desktop, Cursor, and others. The daemon does the thinking (consolidation + extraction); this server is a thin filesystem bridge so agents can read, append, and search memory through the Model Context Protocol.

62
Provisional

Week of 2026-04-27 · Manually assessed

Methodology v2.6.0

Benchmark score

4 dimensions▸

Independent benchmark across four dimensions.

Overall
62.0/100
Reliability
Error handling, retries, and production readiness.
55/100
quality bar: 75 · below benchmark
Latency
Response speed compared to peer agents.
65/100
quality bar: 70 · below benchmark
Cost efficiency
Token cost per successful task.
70/100
category 75th percentile: 78 · below benchmark
Consistency
How dependably the agent completes its stated task, run over run.
68/100
quality bar: 80 · below benchmark

4 improvement opportunities are below their benchmark — sign in to see your ranked fixes.

See your ranked fixes →
  1. ✓complete
    Desk assessment
    Structured multi-model review of published materials.
  2. ○not yet
    Live telemetry
    Connect production telemetry to lift the ceiling from 85 to 97 — 100 once R5 tool-call telemetry lands. Connect telemetry →
  3. How scores upgrade →

Manually assessed by BenchLytix · Week of 2026-04-27

Score reflects an independent capability assessment. Community signals (stars, contributors) appear separately below as adoption indicators that complement — but do not replace — the score.

Security

scan clean▸
Secure

Scan scope: the agent's public GitHub repository. The deployed service may differ from the scanned source.

Last 1 scan

  • 2026-08-16

OWASP MCP Top 10

No current findings.

Runtime sandbox

No current findings.

Supply chain

No current findings.

Community signals

GitHub · 19d▸

Independent adoption indicators from GitHub. These complement — but do not replace — the capability score above.

⭐ Stars
4
👥 Contributors
2
🍴 Forks
0
📂 Open issues
0
🟢 Active· last commit 19 days ago
View on GitHub →

Try Agent Memory

Visit the developer's website to sign up, install, or start a trial.

Visit developer site↗

Reviews