Knowl – CLAUDE.md hit 1000 lines, so I built memory that prunes itself is a skill AI agent for To ensure an AI assistant maintains accurate knowledge over time by automatically discarding outdated information during ongoing conversations. Pricing: not stated on the product page. As of 2026-10-26, KanonAgent records 1 upvotes. First indexed by KanonAgent on 2026-08-22.
Knowl is an AI memory system that automatically removes outdated information from an agent's knowledge base when facts change, keeping long-term interactions accurate.
1 upvotes
Tracked by Kanon since Aug 22, 2026
Status basis: npm_downloads_7d momentum fell to -128 per 3d from 34 (<= 40% of the prior window)
📈 Timelinewhat changed, and when
- Cooling npm_downloads_7d momentum fell to -128 per 3d from 34 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to -128 per 3d from 34 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to 0 per 3d from 592 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to 34 per 3d from 558 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to 34 per 3d from 558 (<= 40% of the prior window)
- Accelerating npm_downloads_7d gained 592 in the last 3d after a flat 3d baseline (no prior movement to compare a ratio against)
- Cooling npm_downloads_7d momentum fell to 34 per 3d from 558 (<= 40% of the prior window)
- Accelerating npm_downloads_7d gained 558 in the last 3d after a flat 3d baseline (no prior movement to compare a ratio against)
- Cooling npm_downloads_7d momentum fell to -235 per 3d from 14 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to -235 per 3d from 55 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to 0 per 3d from 58 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to 14 per 3d from 241 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to 55 per 3d from 148 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to 58 per 3d from 401 (<= 40% of the prior window)
- Cooling npm_downloads_7d momentum fell to 58 per 3d from 401 (<= 40% of the prior window)
- Anomaly daily change -288.0 is -4.3 standard deviations from this entity's own prior mean (10.3±69.2, baseline excludes this point) — unusual, not necessarily adoption
- Anomaly daily change -288.0 is -4.3 standard deviations from this entity's own prior mean (10.3±69.2, baseline excludes this point) — unusual, not necessarily adoption
- Anomaly daily change -288.0 is -4.3 standard deviations from this entity's own prior mean (10.3±69.2, baseline excludes this point) — unusual, not necessarily adoption
- Cooling npm_downloads_7d momentum fell to -288 per 3d from 110 (<= 40% of the prior window)
- Anomaly daily change -288.0 is -4.3 standard deviations from this entity's own prior mean (10.3±69.2, baseline excludes this point) — unusual, not necessarily adoption
- Anomaly daily change -288.0 is -4.3 standard deviations from this entity's own prior mean (10.3±69.2, baseline excludes this point) — unusual, not necessarily adoption
- Cooling npm_downloads_7d momentum fell to -288 per 3d from 110 (<= 40% of the prior window)
- Anomaly daily change -288.0 is -4.3 standard deviations from this entity's own prior mean (10.3±69.2, baseline excludes this point) — unusual, not necessarily adoption
- Anomaly daily change 213.0 is 14.4 standard deviations from this entity's own prior mean (-12.2±15.6, baseline excludes this point) — unusual, not necessarily adoption
🔎 Known / Unknown
11 of 23 fields unknown
| Machine-callable | — | unknown | — |
| Open source | 1 | inferred | 1 quote(s) |
| Self-hostable | 1 | inferred | 1 quote(s) |
| Bring your own key | 1 | inferred | 1 quote(s) |
| Autonomy level | 2 | inferred | 2 quote(s) |
| Pricing model | — | unknown | — |
| Integrations | mcp,claude,cursor,codex | inferred | 1 quote(s) |
“Unknown” means we have not verified it — it is not a “no”. Hard filters never treat unknown as false.
🤖 Agent teardown · skill
Job to be doneTo ensure an AI assistant maintains accurate knowledge over time by automatically discarding outdated information during ongoing conversations.
AutonomyL2 · tool-calling(evidence: "27 MCP tools…")
Who it is forAI developers and teams building long-running AI agents that need reliable, up-to-date memory.
Prerequisitesopen source · self-hostable · bring your own API key
Integrationsmcp · claude · cursor · codex
Traction · why it is risingIt addresses a critical reliability issue in AI agents—persistent memory decay—making it essential for trustworthy, long-term AI interactions.
Why it matters
It solves the core problem of memory staleness in AI agents, making it foundational for building dependable AI assistants.
Evidence quotesverbatim, from the product’s own materials
“public GitHub repository (README fetched)”— structural
“100% local, no egress”— readme
“0 API keys needed”— readme
“27 MCP tools”— readme
“persistent memory across sessions for Claude Code, Cursor and Codex, over MCP”— readme
Signal source: Show HN
Visit official site →
📛 Official badgefor your site / README
Building Knowl? Pick a style above — the embed code updates live. Deep color control via URL params: bg= / fg= / accent= (hex). It links back to this page.
On mobile tap Share for WeChat / RED (Xiaohongshu) / X; on desktop use Copy text and paste into the app.
FAQ
What is Knowl?
Knowl is an AI memory system that automatically removes outdated information from an agent's knowledge base when facts change, keeping long-term interactions accurate.
What does Knowl do?
To ensure an AI assistant maintains accurate knowledge over time by automatically discarding outdated information during ongoing conversations.
Why does Knowl matter?
It solves the core problem of memory staleness in AI agents, making it foundational for building dependable AI assistants.
Is Knowl open source?
Yes — Knowl is open source.
Can I self-host Knowl?
Yes — Knowl can be self-hosted.
What does Knowl integrate with?
mcp,claude,cursor,codex
How popular is Knowl?
As tracked by KanonAgent: 1 upvotes (first indexed 2026-08-22).
What are the best Knowl alternatives?
Similar AI agents tracked by KanonAgent: Stealth Venture, taste-skill, archify, open-design, deer-flow, ruflo.