AWS Lambda Durable Functions
awslabs/agent-plugins
Build resilient, long-running, multi-step applications with AWS Lambda durable functions with automatic state persistence, retry logic, and orchestration for long-running executions.
Gives an agent a workaround when its code-execution sandbox keeps failing: save the Python script to a file and run it through the shell instead.
Install Sandbox Code Execution Fallback
$ npx skills add HKUDS/OpenSpace --skill code-exec-fallback -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install HKUDS/OpenSpace code-exec-fallback --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/HKUDS/OpenSpace.git skills-src && mkdir -p .claude/skills && cp -r skills-src/benchmarks/gdpval/skills/code-exec-fallback .claude/skills/code-exec-fallback && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "code-exec-fallback" agent skill from https://github.com/HKUDS/OpenSpace/tree/main/benchmarks/gdpval/skills/code-exec-fallback into .claude/skills/code-exec-fallback/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "code-exec-fallback", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/HKUDS/OpenSpace/tree/main/benchmarks/gdpval/skills/code-exec-fallbackType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add HKUDS/OpenSpace --skill code-exec-fallback -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install HKUDS/OpenSpace code-exec-fallback --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/HKUDS/OpenSpace.git skills-src && mkdir -p .agents/skills && cp -r skills-src/benchmarks/gdpval/skills/code-exec-fallback .agents/skills/code-exec-fallback && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "code-exec-fallback" agent skill from https://github.com/HKUDS/OpenSpace/tree/main/benchmarks/gdpval/skills/code-exec-fallback into .agents/skills/code-exec-fallback/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "code-exec-fallback", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add HKUDS/OpenSpace --skill code-exec-fallback -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install HKUDS/OpenSpace code-exec-fallback --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/HKUDS/OpenSpace.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/benchmarks/gdpval/skills/code-exec-fallback .cursor/skills/code-exec-fallback && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "code-exec-fallback" agent skill from https://github.com/HKUDS/OpenSpace/tree/main/benchmarks/gdpval/skills/code-exec-fallback into .cursor/skills/code-exec-fallback/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "code-exec-fallback", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/HKUDS/OpenSpace.git --path benchmarks/gdpval/skills/code-exec-fallback--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add HKUDS/OpenSpace --skill code-exec-fallback -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install HKUDS/OpenSpace code-exec-fallback --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/HKUDS/OpenSpace.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/benchmarks/gdpval/skills/code-exec-fallback .gemini/skills/code-exec-fallback && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "code-exec-fallback" agent skill from https://github.com/HKUDS/OpenSpace/tree/main/benchmarks/gdpval/skills/code-exec-fallback into .gemini/skills/code-exec-fallback/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "code-exec-fallback", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install HKUDS/OpenSpace code-exec-fallbackInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add HKUDS/OpenSpace --skill code-exec-fallback -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/HKUDS/OpenSpace.git skills-src && mkdir -p .github/skills && cp -r skills-src/benchmarks/gdpval/skills/code-exec-fallback .github/skills/code-exec-fallback && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "code-exec-fallback" agent skill from https://github.com/HKUDS/OpenSpace/tree/main/benchmarks/gdpval/skills/code-exec-fallback into .github/skills/code-exec-fallback/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "code-exec-fallback", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add HKUDS/OpenSpace --skill code-exec-fallback -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install HKUDS/OpenSpace code-exec-fallback --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/HKUDS/OpenSpace.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/benchmarks/gdpval/skills/code-exec-fallback .opencode/skills/code-exec-fallback && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "code-exec-fallback" agent skill from https://github.com/HKUDS/OpenSpace/tree/main/benchmarks/gdpval/skills/code-exec-fallback into .opencode/skills/code-exec-fallback/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "code-exec-fallback", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
code-exec-fallback- GitHub stars
- 7.7k
- Token cost
- ~588 tokens
- SKILL.md length
- 200 words
- Files
- 2
- Skills in repo
- 199
- Repo updated
- First seen
- Licence
- MIT
At a glance
Gives an agent a workaround when its code-execution sandbox keeps failing: save the Python script to a file and run it through the shell instead.
- Works in 3 steps: Write the Python Script → Execute via Shell → Handle Output
- The code sandbox tool fails twice in a row on the same Python task
- SKILL.md covers When to Use, The Pattern, Step-by-Step Instructions and Example, plus 2 more sections
- Calls python3
What it does
The pattern applies after the execute_code_sandbox tool has failed repeatedly, typically twice or more, because of environment limits, timeouts, missing dependencies or sandbox restrictions. Instead of passing code to the sandbox, the agent saves the Python script with write_file, runs it with run_shell using python3, captures stdout and stderr, and optionally deletes the temporary file afterward.
A few tips keep the shell run reliable: put every import in the script because the shell environment may differ from the sandbox, use absolute paths or confirm the working directory, add error handling, raise the timeout from its 30 second default for long jobs, and print JSON when results must be parsed. It helps most when the sandbox lacks packages, restricts file input and output, times out on work a shell can finish, or when a task needs external commands or a multi-file layout.
When your agent uses it
- The code sandbox tool fails twice in a row on the same Python task
- A script needs packages the sandbox does not have installed
- File reads or writes are blocked inside the sandbox
- A long-running script times out in the sandbox but should work in a shell
Example prompts
- “The sandbox keeps timing out on this CSV analysis, so save it as analyze.py and run it from the shell.”
- “Retry the failed data-cleaning code by writing it to process_data.py and running python3.”
- “The sandbox blocks file writes, so run this report script through the shell and print the result as JSON.”
Requirements
- An agent with write_file and run_shell tools
- Python 3 in the shell environment
Workflow steps
3 steps, taken from the step headings in SKILL.md.
What it can do on your machine
Read from SKILL.md and the folder at commit 3827781. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Shell commands in SKILL.md call:
python3
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Sandbox Code Execution Fallback loads about 588 tokens when it runs. Until then it costs about 23 tokens; SKILL.md has 200 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from HKUDS/OpenSpace at commit 3827781, republished under its MIT licence (© HKUDS). 200 words, ~588 tokens.
Download SKILL.mdSave it as
.claude/skills/code-exec-fallback/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.- name
- code-exec-fallback
- description
- Fallback pattern for executing Python code when execute_code_sandbox fails
Code Execution Fallback
When to Use
Use this pattern when execute_code_sandbox fails repeatedly (typically 2+ attempts) due to environment limitations, timeouts, dependency issues, or sandbox restrictions.
The Pattern
Instead of executing code directly in the sandbox, write the Python script to a file and execute it via shell:
- Write the script using
write_file - Execute via shell using
run_shellwithpython3 script.py - Clean up (optional) remove the temporary file
Step-by-Step Instructions
Step 1: Write the Python Script
Use write_file to save your Python code:
- Path: Choose a descriptive name (e.g., "process_data.py", "analyze.py")
- Content: Your complete Python script with all imports and logicStep 2: Execute via Shell
Use run_shell to execute:
- Command: "python3 <script_name>.py"
- Timeout: Set appropriately for your task (default 30s, increase if needed)Step 3: Handle Output
- Capture stdout/stderr from run_shell
- Parse results as needed
- Optionally delete the script file after executionExample
# Instead of this (which may fail):
execute_code_sandbox(code="import pandas as pd; df = pd.read_csv('data.csv')...")
# Do this:
write_file(path="analyze.py", content="""
import pandas as pd
import json
df = pd.read_csv('data.csv')
result = df.groupby('category').sum()
print(json.dumps(result.to_dict()))
""")
run_shell(command="python3 analyze.py", timeout=60)Tips for Success
- Include all imports in the script file - the shell environment may differ from the sandbox
- Use absolute paths or ensure working directory is correct
- Add error handling to your script for better debugging
- Increase timeout for long-running operations (default is 30s)
- Print structured output (JSON) if you need to parse results
- Clean up temporary files after successful execution to avoid clutter
When This Helps
- Sandbox has missing dependencies
- Code execution times out in sandbox but would work in shell
- File I/O operations are restricted in sandbox
- Need to run external commands or system utilities
- Complex multi-file projects that need proper file structure
© HKUDS, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 1 other file in benchmarks/gdpval/skills/code-exec-fallback of HKUDS/OpenSpace.
- SKILL.md
- .skill_id
Open the folder on GitHubat commit 3827781
Compare with similar skills
Sandbox Code Execution Fallback next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Sandbox Code Execution Fallback this skillHKUDS/OpenSpace | 7.7k | — | ~588 | Automated safety check: Pass | MIT | |
| AWS Lambda Durable Functionsawslabs/agent-plugins | 915 | — | ~2.3k | Automated safety check: Pass | Apache-2.0 | |
| Fake Model Provider Faultsdifferent-ai/openwork | 24k | — | ~642 | Automated safety check: Pass | Custom licence | |
| Cross-Language Coding Standardszereight/gitlab-mcp | 2k | 1 repos | ~1.4k | Automated safety check: Pass | MIT | |
| Python Devdatabricks-solutions/ai-dev-kit | 1.9k | — | ~1.6k | Automated safety check: Pass | Custom licence | |
| Python Script Runnercongchuanling-dot/Cohort | 199 | — | ~577 | Automated safety check: Notes | MIT |
Similar skills
- Official
Fake Model Provider Faults
different-ai/openwork
Makes the desktop app's model provider fail on demand, with refused connections, resets, stalls and HTTP 4xx and 5xx errors, so error and retry states can be reproduced.
Cross-Language Coding Standards
zereight/gitlab-mcp
Shared reference for naming, function size, complexity and error handling rules that reviewer agents apply across TypeScript, Python, Go, Rust, Java, C# and Swift.
Python Dev
databricks-solutions/ai-dev-kit
Python development guidance with code quality standards, error handling, testing practices, and environment management.
Python Script Runner
congchuanling-dot/Cohort
Runs a Python script with checks first: confirms the interpreter version, compiles it for syntax errors, verifies imports, then explains any traceback in plain language.
Robust Error Handling In Scripts
aiming-lab/MetaClaw
A skill your agent uses when writing shell scripts, Python automation, or any unattended batch job.
More from HKUDS/OpenSpace
All 199 skills in this repoAudio Track Production Workflow
HKUDS/OpenSpace
Walks through producing a master audio track plus stems in Python, from checking a reference file and timing sections by BPM to effects, a zip archive and final verification.
Cascade Failure Recovery
HKUDS/OpenSpace
Handle cascading data retrieval tool failures by falling back to embedded knowledge generation
Code Execution Fallback
HKUDS/OpenSpace
A recovery routine for agents whose sandboxed code runner keeps failing: save the Python script to disk, then run it through the shell and read the output.
Code Execution Fallback and Path Anchoring
HKUDS/OpenSpace
Fallback ladder for failed sandboxed code runs, plus the habit of fixing the working directory first so generated files land in the right place.
Code Execution Fallback E81068
HKUDS/OpenSpace
Fallback workflow for executing Python code when executecodesandbox fails repeatedly
Context Anchored Fallback Report
HKUDS/OpenSpace
Generate documents with writefile when retrieval tools fail, with explicit guardrails against task context drift
Related
Works with
Categories
Questions about Sandbox Code Execution Fallback
What does Sandbox Code Execution Fallback do?
Gives an agent a workaround when its code-execution sandbox keeps failing: save the Python script to a file and run it through the shell instead. The pattern applies after the execute_code_sandbox tool has failed repeatedly, typically twice or more, because of environment limits, timeouts, missing dependencies or sandbox restrictions. Instead of passing code to the sandbox, the agent saves the Python script with write_file, runs it with run_shell using python3, captures stdout and stderr, and optionally deletes the temporary file afterward.
When should I use Sandbox Code Execution Fallback?
Sandbox Code Execution Fallback fits situations like: the code sandbox tool fails twice in a row on the same Python task; A script needs packages the sandbox does not have installed; file reads or writes are blocked inside the sandbox; A long-running script times out in the sandbox but should work in a shell.
How do I install Sandbox Code Execution Fallback in Claude Code?
Run `npx skills add HKUDS/OpenSpace --skill code-exec-fallback -a claude-code`. Or copy the skill folder (benchmarks/gdpval/skills/code-exec-fallback in HKUDS/OpenSpace) into .claude/skills/code-exec-fallback in your project. Claude Code loads it when a task matches its description.
How do I install Sandbox Code Execution Fallback in Codex?
Run `npx skills add HKUDS/OpenSpace --skill code-exec-fallback -a codex`. Or copy the skill folder (benchmarks/gdpval/skills/code-exec-fallback in HKUDS/OpenSpace) into .agents/skills/code-exec-fallback in your project. Codex loads it when a task matches its description.
Can I use Sandbox Code Execution Fallback in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add HKUDS/OpenSpace --skill code-exec-fallback -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/code-exec-fallback, .gemini/skills/code-exec-fallback, .github/skills/code-exec-fallback and .opencode/skills/code-exec-fallback in your project.
What does Sandbox Code Execution Fallback need to run?
Going by SKILL.md and its folder, Sandbox Code Execution Fallback needs the command-line tools its instructions call (python3). Our summary lists: An agent with write_file and run_shell tools; Python 3 in the shell environment.
Does Sandbox Code Execution Fallback access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is Sandbox Code Execution Fallback safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Sandbox Code Execution Fallback use?
Sandbox Code Execution Fallback is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Sandbox Code Execution Fallback use?
About 588 tokens (SKILL.md is roughly 2.4k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Sandbox Code Execution Fallback?
Skills that share tags, products or a category with Sandbox Code Execution Fallback: AWS Lambda Durable Functions (awslabs/agent-plugins, 915 stars), Fake Model Provider Faults (different-ai/openwork, 24k stars), Cross-Language Coding Standards (zereight/gitlab-mcp, 2k stars) and Python Dev (databricks-solutions/ai-dev-kit, 1.9k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Sandbox Code Execution Fallback?
HKUDS (a GitHub organization) maintains it in HKUDS/OpenSpace, which has 7,749 GitHub stars. The repository holds 199 skills in this directory. The repository was last updated on August 12, 2026.
Source: HKUDS/OpenSpace on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.