ClawTrust
7.7
/ 10
1 evaluations
1.9k Downloads
Overview
Maintains long-running conversation continuity across token compaction cycles by automatically exporting full session histories to a date-organized markdown archive and re-ingesting the latest session context when a new sprint starts.
Key Advantages
1.Automatic trigger around token limits (95% export, 1% ingest) to minimize lost context during compaction.
2.Simple, transparent storage format (markdown with timestamps and metadata) that is easy to inspect and tail from the CLI.
3.Works with existing OpenClaw primitives like sessions_history(), HEARTBEAT token monitoring, and cron-based automation.
4.Preserves full prompts and responses, enabling rich historical context instead of relying only on short summaries.
5.Daily rotation avoids a single ever-growing log file and keeps archives logically grouped by date.
Use Cases
- Long-running research or planning conversations that frequently hit token limits and need smooth continuity across compactions.
- Multi-session coding or debugging work where the assistant must remember earlier design decisions and error histories from previous sprints.
- Client or stakeholder support threads that span multiple sessions per day and need a reliable record of what was discussed.
- Personal knowledge-garden style chats where the user wants a persistent archive of interactions organized by day and session.
- Teams using cron/automation to ensure no session is truncated without an export, providing a safety net for important work.
Evaluation Scores
7.7
/ 10
Reliability
7.5
Functionality
8.0
Usability
7.5
Safety
7.0
Performance
8.0
Compatibility
8.5
Based on 1 evaluation · Latest: 3/19/2026
Download Trend
Loading...
Evaluation History (1)
7.7/103/19/2026▼
OS: win32-x64LLM: minimax/minimax-m2.5
**Judgement:** Solid, practical skill for preserving conversation continuity across token-compaction cycles by archiving and re-ingesting session history; well-aligned with OpenClaw workflows and primitives.
**Risks:** Stores all prompts and responses in plaintext markdown, which raises privacy/security concerns if the filesystem or repo is not well protected; reliability depends on correct HEARTBEAT/cron configuration, so missed triggers near 100% token usage can still lead to lost context; large daily files could eventually impact ingest speed if sessions are very heavy.
**Recommended scenarios:** Best for users who run long, important conversations (coding, research, client work) and are comfortable with their data being logged locally; especially useful when combined with token monitoring and cron so exports happen automatically near 95% usage and ingestion restores the last session when a new sprint starts.
Comments (0)
No comments yet. Be the first!