All Claude Code modskzarzycki/claude-mods
eval-kernel
A persistent JavaScript kernel the model drives with one `eval` tool: state survives between cells, and cells call Claude Code tools, models and subagents.
eval-kernel reaches L3 (network) by its classified calls and passed claude plugin validate with warnings on Claude Code 2.1.302.
- Events hooked: 5
- API methods referenced: 13
- Access from classified calls: L3, network: network, drives Claude, draws
- Unclassified calls: 3 (
$.decision.ask,$.process.spawn,$.tool.list), full access unknown - Static scan, not a runtime test
Install hint
/plugin install eval-kernel --marketplace kzarzycki/claude-modsFrom the repository's own marketplace. Confirm the plugin name in the repository's README before installing. Mods need Claude Code 2.1.287 or later.
Access
L3 Uses the networkother: $.decision.ask, other: $.process.spawn, other: $.tool.list, network, drives Claude, drawsValidation
Passed with warnings on Claude Code 2.1.302View the source on GitHub Back to the list
What the scan recorded
- Observed events
- mcp__eval-kernel__eval calls, every tool call, every prompt
- Hooks
session.start, tool.call {tool=mcp__eval-kernel__eval}, tool.call, prompt.submit, turn.complete, command.run {command=eval}- API calls
$.agent.spawn, $.clock.sleep, $.command.register, $.decision.ask, $.http.fetch, $.model.complete, $.process.spawn, $.session.cwd, $.session.id, $.tool.call, $.tool.list, $.tool.register, $.ui.log- Validation on 2.1.302
- Passed with warnings
- Scanned revision
bc9397750972- Marketplace warnings
- claude-mods
These are the hooks and API calls reported by claude plugin validate. Access levels describe the widest reach of those calls and are not safety ratings. Validation is a static check, not a runtime test, and a passing result does not prove that a mod works or is safe. How the scan works.
More from kzarzycki/claude-mods
- agent-hubAgent hub: a pane listing this session's subagents with status, model, turns and tokens, and a box to steer a…
- auto-effortAuto effort: the decision model rates each prompt and the turn it starts runs at that effort.
- compact-methodsoh-my-pi compaction methods: shake (move old tool output to files, no model call) and handoff (continue from…
- decision-modelA decision model for other mods ($.decision.ask): typed questions about a state, answered by any System One e…
- resume-on-stopUnexpected stops: when a turn ends right after saying it will act, the decision model spots it and one turn r…
- self-compactCompaction at task boundaries: the model can ask to compact once its turn ends. Above a fill threshold it is…
- ttsr-rulesoh-my-pi TTSR rules: a regex rule denies the tool call that would break it, with the rule as the reason; a qu…
Badge for your README
Authors can show the latest scan result in their README. Each badge links to this page.
[](https://mods.aidojo.si/mods/kzarzycki--claude-mods--eval-kernel/)
[](https://mods.aidojo.si/mods/kzarzycki--claude-mods--eval-kernel/)