Skip to content

donaldfilimon/abi / .agents/skills/os-control-dryrun

os-control-dryrun

Verified

Build the abi CLI and exercise the OS-control policy in dry-run (planning) mode only, and verify the execute path refuses without --confirm. Use when working on OS-control handlers or the `abi agent os` handler, or to safely demo OS-command planning. Never executes a command.

os-control-dryrun — safely drive abi's OS-control planning

Driver: .agents/skills/os-control-dryrun/dryrun.sh (paths relative to repo root). Read-only-effect CLI check — evidence is the RESULT: line. Nothing is executed.

Run (agent path)

.agents/skills/os-control-dryrun/dryrun.sh                 # defaults to: pwd
.agents/skills/os-control-dryrun/dryrun.sh ls -la           # approved option argv

Builds the CLI, runs abi agent os dry-run <command> [args...] (asserts the dry-run: marker), then asserts abi agent os execute without --confirm is refused with usage (exit 2). The allow-list accepts command argv, not narrative text; ls permits approved read-only options but no filesystem operands. Prints RESULT: PASS (exit 0) or a FAIL count.

Current Rust driver: proves dry-run emits a plan and execute without --confirm exits 2; it never invokes the confirmed execution path.

Gotchas

  • ⚠️ This skill never runs agent os execute --confirm — that actually runs the planned OS command and is deliberately out of scope. Only dry-run (plan) and the negative confirm-gate check are exercised.
  • The compiled ceiling and narrow-only policy live under crates/abi-cli/src/os/. Use the os-control-policy-reviewer subagent to audit the validation/whitelist.

Troubleshooting

SymptomFix
build FAIL./tools/check.sh.
execute-without-confirm didn't return 2the confirm gate regressed — check crates/abi-cli/src/agent.rs agent-os path.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall