From context-surgeon
Use this skill when configuring, updating, or operating Consurg for agent context control (scopes, permission tiers, wire integrations, tracing, and enforcement) in CLI workflows.
How this skill is triggered — by the user, by Claude, or both
Slash command
/context-surgeon:context-surgeonThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill for any request to set up or adjust Consurg as a context-control tool for AI coding assistants.
Use this skill for any request to set up or adjust Consurg as a context-control tool for AI coding assistants.
working_set (Tier 4): read/write files.reference (Tier 3): read-only files.signatures (Tier 2): API/type signatures.visible (Tier 1): file existence only.consurg init <scope-name>
consurg add src/path.py tests/test_path.py
consurg add --read pyproject.toml docs/notes.md
consurg add --sig src/types.py
consurg on
consurg status
consurg trace <entry-file> --depth <N>
consurg on
consurg git-diff [ref]
consurg on
consurg off
consurg remove src/path.py
consurg clean
Connect Consurg enforcement to an assistant tool:
consurg wire claudeconsurg wire codexconsurg wire geminiconsurg wire pk-agentconsurg wire droidUse --unwire to remove generated hook or integration files.
consurg guard starts interactive runtime guard.consurg wrap -- <command> runs a command inside Consurg enforcement.consurg pin / consurg wire <tool> --unwire handle durable scope/session state.consurg export --format <claude|cursor|aider|generic> exports a scope.consurg map and consurg status inspect active state.working_set, then add reference/signatures as needed.explorer: true only when needed, then remove it before final edits.npx claudepluginhub kingkillery/consurgConfigures tool permissions for Claude Code agents via --allowedTools flags, path patterns for file access, and Bash command prefixes. Reference for secure launches.
Configures Claude Code permissions: tool rules (allow/deny/ask), modes (plan/dontAsk/bypass), sandboxing. Use for Bash/Edit/WebFetch policies, debugging prompts, org managed settings.
Provides reference for Claude Code permission modes (default, acceptEdits, plan, dontAsk, bypass), allow/deny lists, pattern matching, and tool categories. Use to configure secure tool access and switch modes runtime.