How do I install the Codex Local Sdk Usage skill?

Run npx skillmds@latest add maestromaximo/codex-local-sdk-usage in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Codex Local Sdk Usage safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: executes scripts. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Codex Local Sdk Usage?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Codex Local Sdk Usage free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Codex Local Sdk Usage?

maestromaximo (@maestromaximo) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Codex Local Sdk Usage by maestromaximo

Name: Codex Local Sdk Usage
Rating: 4.2 (5 reviews)
Author: maestromaximo

Codex Local Sdk Usage

Use when Codex needs to write, wire, or troubleshoot Python code that consumes this repository's `codex_local_sdk` package (not modify SDK internals). Trigger for requests to call `CodexLocalClient`, run or resume threads, stream JSON events, configure retries and timeouts, use schema-constrained output, or persist named sessions.

maestromaximo Updated Sep 10, 2026

File contents

Codex Local SDK Usage

Route Quickly

Load only the reference file needed for the task:

Select an execution path: references/task-routing.md
Check methods, request flags, and result fields: references/sdk-api-cheatsheet.md
Debug failures, retries, and session issues: references/guardrails-and-failure-modes.md

Execute Workflow

Run scripts/preflight_check.py to verify Python, package import, and codex CLI presence.
Run scripts/scaffold_usage_template.py --list and choose the closest template.
Run the scaffold command with --template and --output.
Replace placeholders (prompt, schema, paths, session names, retries) and keep signatures aligned with the API cheat sheet.
Run the generated script and inspect CodexExecResult (return_code, turn_status, stderr, final_message).
If execution fails, apply the guardrails reference and retry.

Use Bundled Scripts

scripts/preflight_check.py: print readiness checks with optional strict mode.
scripts/scaffold_usage_template.py: copy a starter script from assets/.

Use Bundled Templates

assets/template_sync_run.py
assets/template_live_stream.py
assets/template_thread_session.py
assets/template_resume_named_session.py
assets/template_async_run.py
assets/template_schema_output.py
assets/template_telemetry_and_retry.py

Keep Consumer Guardrails

Set timeout_seconds for run* and resume* unless intentionally unbounded.
Set json_output=True whenever you need structured events, thread_id, or turn_status.
Catch CodexExecFailedError when raise_on_error=True; inspect exc.result.
Pass session_id or session_name, never both.
Treat live methods (run_live*, resume_live*) as startup-retry only; once a handle is returned, caller code owns interruption/retry logic.
On Windows, if you hit CodexError: ... [WinError 2], initialize with CodexLocalClient(codex_bin="codex.cmd").

---name: codex-local-sdk-usagedescription: Use when Codex needs to write, wire, or troubleshoot Python code that consumes this repository's `codex_local_sdk` package (not modify SDK internals). Trigger for requests to call `CodexLocalClient`, run or resume threads, stream JSON events, configure retries and timeouts, use schema-constrained output, or persist named sessions.---# Codex Local SDK Usage## Route QuicklyLoad only the reference file needed for the task:- Select an execution path: `references/task-routing.md`- Check methods, request flags, and result fields: `references/sdk-api-cheatsheet.md`- Debug failures, retries, and session issues: `references/guardrails-and-failure-modes.md`## Execute Workflow1. Run `scripts/preflight_check.py` to verify Python, package import, and `codex` CLI presence.2. Run `scripts/scaffold_usage_template.py --list` and choose the closest template.3. Run the scaffold command with `--template` and `--output`.4. Replace placeholders (`prompt`, schema, paths, session names, retries) and keep signatures aligned with the API cheat sheet.5. Run the generated script and inspect `CodexExecResult` (`return_code`, `turn_status`, `stderr`, `final_message`).6. If execution fails, apply the guardrails reference and retry.## Use Bundled Scripts- `scripts/preflight_check.py`: print readiness checks with optional strict mode.- `scripts/scaffold_usage_template.py`: copy a starter script from `assets/`.## Use Bundled Templates- `assets/template_sync_run.py`- `assets/template_live_stream.py`- `assets/template_thread_session.py`- `assets/template_resume_named_session.py`- `assets/template_async_run.py`- `assets/template_schema_output.py`- `assets/template_telemetry_and_retry.py`## Keep Consumer Guardrails- Set `timeout_seconds` for `run*` and `resume*` unless intentionally unbounded.- Set `json_output=True` whenever you need structured events, `thread_id`, or `turn_status`.- Catch `CodexExecFailedError` when `raise_on_error=True`; inspect `exc.result`.- Pass `session_id` or `session_name`, never both.- Treat live methods (`run_live*`, `resume_live*`) as startup-retry only; once a handle is returned, caller code owns interruption/retry logic.- On Windows, if you hit `CodexError: ... [WinError 2]`, initialize with `CodexLocalClient(codex_bin="codex.cmd")`.

Frequently asked questions