Back to skill

Security audit

Subagent Driven Development

Security checks for vulnerabilities and agentic risk

Overview

This skill automates development work in a Git workspace, with disclosed local mutation and cleanup behavior that is mostly scoped to the stated purpose.

Install only if you want an agent to make code changes and commits from an existing plan with limited check-ins. Before final cleanup, require the agent to print and verify that the workspace path is the intended per-plan .superpowers/sdd directory.

Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill instructs deletion of a workspace with rm -rf <workspace> after completion, but does not require an explicit confirmation step at the point of destruction or strong path-safety validation. In an agentic setting, destructive filesystem actions are risky because path resolution mistakes, variable corruption, or misuse of the workspace helper could delete unintended data.

Content

No source excerpt is available for this finding.

Tool Parameter Abuse

High
Category
Tool Misuse
Confidence
97% confidence
Finding

The skill explicitly references git clean -fdx, a highly destructive command that recursively deletes all untracked and ignored files in the repository. Even though it is framed as a warning about what could happen, including such a command in an agent skill materially increases the risk of unsafe execution, especially if reused by an autonomous agent or copied into operational flow without safeguards.

Content

Scanner excerpt · SKILL.md (reported line 231)May include surrounding context.

md
- The ledger is your recovery map: the commits it names exist in git even
  when your context no longer remembers creating them. After compaction,
  trust the ledger and `git log` over your own recollection.
- `git clean -fdx` will destroy the workspace (it's git-ignored scratch); if
  that happens, recover from `git log`.

Read the plan once, note its context and Global Constraints, and create a

YARA rule 'agent_skill_destructive_autonomous_actions': Autonomous destructive filesystem, shell history, or repository actions in AI agent skills [agent_skills]

High
Category
YARA Match
Confidence
90% confidence
Finding

The combined pattern is supported by the skill's overall posture: it promotes continuous autonomous execution, includes silent/default behaviors, and references destructive cleanup behavior. In a development-agent context, chaining autonomous task execution with destructive filesystem operations raises the chance of accidental data loss or unintended repository modification without a clear user checkpoint.

Content

Scanner excerpt · SKILL.md (reported line 232)May include surrounding context.

md
when your context no longer remembers creating them. After compaction,
  trust the ledger and `git log` over your own recollection.
- `git clean -fdx` will destroy the workspace (it's git-ignored scratch); if
  that happens, recover from `git log`.

Read the plan once, note its context and Global Constraints, and create a
todo per task.