TPM Copilot

AI-powered operating system for Technical Program Managers and Project Managers. Pulls data from Jira, Linear, GitHub, and calendars to auto-generate status...

This package is internally consistent with its claimed purpose, but you should: (1) review and place API credentials deliberately — use least-privilege tokens (e.g., project-scoped Jira tokens, machine/service accounts where possible), (2) inspect the generated workspace ($HOME/.openclaw/workspace/tpm by default) and config.json before running, (3) be aware scripts can create Jira issues and post to Slack/email — test with a sandbox project/webhook first, (4) avoid supplying org-wide admin tokens: prefer individual or service-account tokens with limited scopes, (5) verify gh CLI is authenticated to the correct GitHub account and test gh commands manually, and (6) if you need the registry to reflect required env vars, ask the publisher to update metadata so automated permission checks can be accurate.

Type: OpenClaw Skill Name: tpm-copilot Version: 1.0.0 The skill is classified as suspicious due to critical vulnerabilities that could lead to remote code execution (RCE) and arbitrary file system manipulation. Specifically, `scripts/add-program.sh` is vulnerable to path traversal and command injection via the `--name` argument, allowing an attacker to create files/directories outside the intended workspace or execute arbitrary commands. Furthermore, `scripts/risk-radar.sh` and `scripts/status-report.sh` execute `gh` CLI commands using `subprocess.run` where the `--repo` argument is sourced from `programs/<name>/config.json`. If an attacker can manipulate this configuration file (potentially by exploiting the `add-program.sh` vulnerability), they could inject shell commands into the `repo` field, leading to RCE. The skill also handles sensitive API keys (Jira, Linear, GitHub, Slack, Resend), making it a high-value target if these vulnerabilities are exploited.