Testing claude-mem for persistent memory in coding agents
11/14/2026 — 11/14, 13:17·1 sources (all)·1 reports (all)
Key takeaways
- The author installed claude-mem, an Apache-2.0 project with 116k GitHub stars, and ran the same task in fresh sessions with memory on and of…
Background & analysis
On November 14, 2026, the AI section of DEV Community published a hands-on article titled "Testing claude-mem: Adding Persistent Memory to AI Coding Agents." A few weeks earlier, the author had tested billion-context, a tool that compresses a long coding-agent session so it fits inside a small context window, but that approach only addresses a single session. The other half of the problem, the author notes, surfaces the following morning: when a brand-new session is opened, the agent has forgotten everything, including decisions the two sides made together the day before. claude-mem targets precisely that second half.
According to the article, claude-mem is released under the Apache-2.0 license, has accumulated 116k stars on GitHub, and has been in existence since August 2025. Several explainers about it had already appeared on Medium, so the author did not repeat those overviews. To measure what the tool actually does, the author installed claude-mem and ran the same task in a fresh session twice, once with memory enabled and once with it disabled. The comparison tracked the amount of context per request, the time taken, the number of tool calls, and how often the agent had to look up information it should already know, offering a practical check of persistent memory for coding agents.
As of this report, the available material contains no further developments on this particular test, leaving the evaluation at the point of the author's measured comparison.
AI-generated from 1 reports · updated 3 days ago
Related AI tools
- Heat index
- 25
- All sources
- 1
- All reports
- 1
- First seen
- 3 days ago
Reports on this story headlines open the original
The author installed claude-mem (Apache-2.0, 116k GitHub stars) and ran the same task in fresh sessions with and without it, comparing context per request, wall time, tool calls, and how often the agent looked up known information.
DEV Community · AIAI score 53
Other stories people are talking about
- Heat index 80Developer measures how screenshot resolution drives Claude vision token cost1 sources
- Heat index 90NewDeveloper benchmarks Claude Code plugins across four tech stacks1 sources
- Heat index 47Testing three Claude code-review skills on 1,306 comments1 sources
- Heat index 80Open-source Agent Keychain lets agents use credentials without seeing them1 sources
- Heat index 108Claude Code Code Mode can be enabled via an environment variable2 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

