Skills/PyPI/mcp-colab-gpu

mcp-colab-gpu@0.3.1

CAUTION· not started

Publisher claims: "MCP server for executing Python on Google Colab GPU/TPU runtimes — supports all Colab Pro accelerators (T4/L4/A100/H100/G4) with high-memory option and security hardening. Fork of mcp-server-colab-exec by Paritosh Dwivedi." Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry PyPI (MCP server)Publisher mio-github · publishing for 248 daysInstalls 20Last detonated 7d ago
What would fix thisWatchDispute verdict

Coverage

Not started: dependency install failed; the server itself was not observed.
Tested
the dependency install step · every file read, file written and host contacted by what ran
Not tested
server startup · the tool list · tool behavior

What this means for the label: Caution here covers what ran. The server's tools were not exercised with a real credential, so nothing about their behavior is asserted; the credential summary below says what to do with that.

Before you give it your token

Code inventory arrives with this version's next honeypot run.
It will ask for
unknown until the next run
Where it talks to
unknown until the next run
What the honeypot saw
no credential files read · no network
What we could not test
server startup · the tool list · tool behavior

Not enough was observed to say yet. Not started: dependency install failed; the server itself was not observed. This version's code has not been inventoried for credentials and hosts either; both arrive with its next honeypot run (popular servers first). `vet-skills try mcp-colab-gpu` runs it now and prints the full answer.

Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.

Is [email protected] safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (2 scenarios; not started: dependency install failed; the server itself was not observed) on 2026-11-07 and scored it 72/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Destructive file system command detected; Destructive command against a filesystem root or device; Destructive command against a filesystem root or device. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 0 s · 2 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
0.1sprocessdependency install exit 1
0.1sprocessexec /usr/bin/date
0.1sprocessexec /usr/bin/timeout
0.1sprocessexec .venv/bin/mcp-colab-gpu
0.1sclaimedno tools/list reply (exit 0)
0.1sprocessscript backend runs setup only
0.1sprocessno install scripts to re-run

Static findings

MEDIUM
File permission modification command.claude/settings.local.json:16 · File permission modification command — agent changes file ownership/permissions"Bash(chmod +x /Users/masayahirano/.claude/run-colab-gpu.sh)",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
HIGH
Destructive file system command detected.claude/settings.local.json:20 · Destructive file system command detected"Bash(rm -rf \"/Volumes/KIOXIA 1TB/Developments/withAI/Vercel/MioSystemCo.Ltd./mcp-colab-gpu-repo/.venv\" 2>/dev/null; rm -rf \"/Volumes/KIOXIA 1TB/Developments/withAI/Vercel/MioSystemCo.Ltd./mcp-colaFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
CRITICAL
Destructive command against a filesystem root or device.claude/settings.local.json:20 · Destructive command against a filesystem root or device"Bash(rm -rf \"/Volumes/KIOXIA 1TB/Developments/withAI/Vercel/MioSystemCo.Ltd./mcp-colab-gpu-repo/.venv\" 2>/dev/null; rm -rf \"/Volumes/KIOXIA 1TB/Developments/withAI/Vercel/MioSystemCo.Ltd./mcp-colaFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
File permission modification command.claude/worktrees/agent-aa3a4360/.claude/settings.local.json:16 · File permission modification command — agent changes file ownership/permissions"Bash(chmod +x /Users/masayahirano/.claude/run-colab-gpu.sh)",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
CRITICAL
Destructive command against a filesystem root or device.claude/worktrees/agent-aa3a4360/.claude/settings.local.json:20 · Destructive command against a filesystem root or device"Bash(rm -rf \"/Volumes/KIOXIA 1TB/Developments/withAI/Vercel/MioSystemCo.Ltd./mcp-colab-gpu-repo/.venv\" 2>/dev/null; rm -rf \"/Volumes/KIOXIA 1TB/Developments/withAI/Vercel/MioSystemCo.Ltd./mcp-colaFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Destructive file system command detected.claude/worktrees/agent-aa3a4360/.claude/settings.local.json:20 · Destructive file system command detected"Bash(rm -rf \"/Volumes/KIOXIA 1TB/Developments/withAI/Vercel/MioSystemCo.Ltd./mcp-colab-gpu-repo/.venv\" 2>/dev/null; rm -rf \"/Volumes/KIOXIA 1TB/Developments/withAI/Vercel/MioSystemCo.Ltd./mcp-colaFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
File type combined with upload actiontests/e2e_drive_test.py:114 · File type combined with upload action — agent may exfiltrate local data filesf"{DRIVE_TEST_FOLDER}/input.csv": "/content/input.csv",Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
File type combined with upload actiontests/e2e_drive_test.py:101 · File type combined with upload action — agent may exfiltrate local data filesmode="w", suffix=".csv", delete=False, prefix="mcp_e2e_input_"Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
File type combined with upload actiontests/e2e_drive_refresh_test.py:118 · File type combined with upload action — agent may exfiltrate local data filessuffix=".csv", delete=False, prefix="mcp_e2e_dl_input_"Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
File type combined with upload actiontests/e2e_drive_refresh_test.py:50 · File type combined with upload action — agent may exfiltrate local data filesmode="w", suffix=".csv", delete=False, prefix="mcp_e2e_input_"Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
File type combined with upload actiontests/e2e_drive_test.py:117 · File type combined with upload action — agent may exfiltrate local data files"/content/output.csv": f"{DRIVE_TEST_FOLDER}/output.csv",Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
Access to sensitive file pathstests/test_drive.py:241 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesdownload_from_drive("data/file.csv", "/tmp/../../etc/passwd", mock_creds)Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
File type combined with upload actiontests/e2e_drive_test.py:180 · File type combined with upload action — agent may exfiltrate local data filesf"{DRIVE_TEST_FOLDER}/output.csv", output_path, credsFix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
Access to sensitive file pathstests/test_drive.py:63 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files_validate_local_path("/tmp/../../etc/passwd")Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
File type combined with upload actiontests/e2e_drive_test.py:173 · File type combined with upload action — agent may exfiltrate local data filessuffix=".csv", delete=False, prefix="mcp_e2e_output_"Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07

Cost to the agent

Instructions loaded
≈92 tok
1 file · 31 in bundle
Tool list per prompt
≈0 tok
0 tools · 0 per tool
Handshake
no reply
launch to tools/list, stdio
Undocumented tools
0
description under 20 chars · avg 0

Tokens estimated at four characters each. The tool list is re-read by the agent on every turn, so it is the number to compare between servers that do the same job: ranked across the index.

Known vulnerabilities

No advisories in OSV for mcp-colab-gpu@0.3.1 as of 2 h ago. This is the package itself, not its dependencies.

What would make this Verified

8 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%
  1. Destructive command against a filesystem root or device≈ +9 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
  2. Unclaimed publisher≈ +8 on the totalClaim mio-github: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
  3. Destructive file system command detected≈ +4 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
  4. File permission modification command≈ +2 on the totalRequest the narrowest permission that works; replace wildcards with the exact paths or hosts.
  5. File type combined with upload action≈ +2 on the totalRemove the call, or declare the host in frontmatter and explain in the README what is sent and why.
  6. Access to sensitive file paths≈ +2 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
Publish a new version to PyPI, then queue it here. Not your skill? Claim the handle to get verdict changes by email.Claim mio-githubPropose a patch to the maintainer