ClawHub

OpenClaw Context Upgrade

v1.0.0

Diagnose and upgrade OpenClaw conversation context limits, especially moving a chat from ~272k to 1M. Use when the user asks to enlarge context, make GPT-5.4...

⭐ 0· 202·0 current·0 all-time
by@gyliiiiii
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
✓
Purpose & Capability
The name/description (increase conversation context to ~1M) aligns with the runtime instructions (inspect agents.defaults.contextTokens, pin model, config.patch, restart, verify). No unrelated credentials, binaries, or external services are requested.
✓
Instruction Scope
SKILL.md stays on-task: it tells the agent to read session status, query config.schema.lookup, patch agents.defaults.contextTokens, restart, start a fresh session, and re-verify. It does not instruct reading unrelated files or exfiltrating data. Note: it assumes gateway config.patch will restart OpenClaw automatically—confirm this behavior on your platform.
✓
Install Mechanism
No install spec and no code files (instruction-only), so nothing is written to disk or downloaded during install.
✓
Credentials
No environment variables, credentials, or config paths are required beyond the OpenClaw config keys the skill legitimately manipulates (agents.defaults.*). The requested access is proportional to changing default context limits.
✓
Persistence & Privilege
always is false and the skill does not request elevated or persistent privileges. It does instruct making configuration changes that will persist in OpenClaw and trigger a restart—this is appropriate for the stated goal but requires authorization by the platform operator.
Assessment
This skill is coherent and low-risk in terms of installs or secrets, but it performs configuration changes and triggers a service restart. Only run it if you (or your admin) authorize modifying OpenClaw defaults. Before applying changes: (1) ask the skill to only diagnose and show intended patches (dry-run), (2) confirm GPT-5.4 is available from your provider, (3) back up current config, and (4) require explicit user/admin confirmation before applying the config.patch that restarts the service. If verification still shows a lower cap, expect provider/model limits to be the blocker rather than OpenClaw settings.

Like a lobster shell, security has layers — review code before you run it.

latestvk97e35y69mp45316qjgfa24e6582ta0t

License

MIT-0
Free to use, modify, and redistribute. No attribution required.
Termshttps://spdx.org/licenses/MIT-0.html

Comments