agent-graph

What the scan recorded

Observed events

every tool call, Agent calls, mcp__agent-graph__read_graph calls, mcp__agent-graph__edit_graph calls, subagent spawns, classic hooks

Hooks

session.start, command.run {command=agent-graph}, ui.close {id=agent-graph}, ui.render {component=AbovePrompt}, ui.render {component=Pane, requestId=agent-graph}, ui.message {element=canvas}, tool.call, tool.call {tool=Agent}, tool.call {tool=mcp__agent-graph__read_graph}, tool.call {tool=mcp__agent-graph__edit_graph}, agent.spawn, turn.step, classic.StopFailure, turn.complete

API calls

$.agent.list, $.agent.register, $.agent.spawn, $.clock.after, $.clock.now, $.command.register, $.env.get, $.fs.exists, $.fs.list, $.fs.read, $.fs.stat, $.fs.write, $.http.fetch, $.model.complete, $.process.run, $.process.spawn, $.prompt.fill, $.prompt.submit, $.session.authorize, $.state.get, $.state.set, $.store.delete, $.store.get, $.store.set, $.tool.call, $.tool.check, $.tool.list, $.tool.register, $.ui.ask, $.ui.close, $.ui.open, $.ui.resolve, $.ui.toast

Surface modules

hooks/canvas.tsx

Validation on 2.1.304

Passed with warnings

Scanned revision

de5c10aba20a

UI rewrite review

UI hook source contains control-character strings. Review any values passed to next() as rewritten props.text; the scanner does not trace whether these strings reach that call. agent-graph/hooks/actions.ts:486, agent-graph/hooks/cloud/api.ts:11, agent-graph/hooks/cloud/api.ts:12, agent-graph/hooks/cloud/api.ts:14, agent-graph/hooks/cloud/backend.ts:101

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. What a mod can access and how the scan works.