メニュー
SWE-Agent inspired coding workflow for fixing, reviewing, and improving software repositories through a compact agent-computer interface.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
Register AgentImage packages, publish checkpoint-derived images, or replace the current process image. Use when creating reusable runtime configurations or deliberately changing process code, tools, prompt identity, memory, or authority.
Diagnose this process's Capability authority and request one exact Human policy decision when a protected operation lacks requestable authority. Do not use it for intent questions, validation errors, provider failures, or speculative privilege expansion.
Delegate one attenuated Capability to a known direct child or revoke one exact Capability. Do not create children, grant peers or grandchildren, amplify authority, or treat revocation as rollback of completed effects.
Capture, inspect, compare, restore, or fork durable process-subtree checkpoints. Use for recoverable internal-state milestones, isolated replay, or deliberate rollback of reconstructable Agent libOS state.
Spawn or fork direct AgentProcess children, coordinate their lifecycle and messages, then import selected results. Use for bounded parallel delegation, isolated exploration, or structured parent-child collaboration.
Run one policy-governed, non-interactive host command through the argv-only Shell boundary. Use for builds, tests, searches, or utilities lacking a narrower tool; not for general Git, TTY/streaming work, ad hoc remotes, or attempts to evade policy.
SOC 職業分類に基づく
| name | swe-agent |
| description | SWE-Agent inspired coding workflow for fixing, reviewing, and improving software repositories through a compact agent-computer interface. |
| license | Apache-2.0 |
| compatibility | agent-libos>=0.1.0 |
| allowed-tools | ["read_directory","read_text_file","write_text_file","write_directory","run_shell_command","get_working_directory","set_working_directory","parse_pytest_log","create_checkpoint","diff_checkpoint","create_memory_object","append_memory_object","read_memory_object","create_object_from_file","write_object_to_file","request_permission","human_output","process_exit"] |
| metadata | {"agent-libos.version":"v0","agent-libos.actions":"references/agent-libos/actions.json","agent-libos.required-capabilities":"references/agent-libos/required-capabilities.json","agent-libos.jit-tools":"references/agent-libos/jit-tools.json"} |
Use this skill when the goal is to fix, review, or improve a software repository in a SWE-Agent style loop.
swe_view for directory listings and bounded file windows with line numbers.swe_grep for concise repository search.swe_edit for exact-text or line-range edits.swe_run for tests and diagnostics.swe_submit when the issue is resolved and evidence is ready.swe_edit intentionally refuses sources whose bounded 1 MiB read is
truncated; use a complete-file-safe editing workflow for larger files.swe_run caps its mediated shell deadline at 55 seconds inside a 60-second
Deno sandbox window; swe_grep uses a 10-second shell deadline inside a
15-second sandbox window. The outer margin is reserved for returning the
observation and cleaning up the contained Deno process. swe_submit stores its
status, summary, tests, and residual risks as the process-exit result payload.
This skill reproduces the useful SWE-Agent ACI shape inside Agent libOS. It does not grant filesystem, shell, process, human, object, or remote authority; every side effect still goes through libOS primitives and process capabilities.