ClawHub
Back to skill

Security audit

fitctl

Security checks for vulnerabilities and agentic risk

Overview

This skill appears benign: it is a disclosed local Linux hardware-inspection and fit-validation helper with optional pinned example downloads and workspace report/cache writes.

Install only if you are comfortable with local hardware/resource evidence collection and JSON reports being written in your workspace. Use the isolated venv/cargo roots described by the skill, run example scripts only when GitHub access and a local cache are acceptable, inspect or redact artifacts before sharing them, and explicitly authorize any path probes that write temporary files.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (9)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The skill is presented primarily as local Linux hardware inspection and fit validation, but the document also includes workflows that fetch replay inputs from GitHub and maintain a local cache. This behavior mismatch is dangerous because operators may approve the skill for local-only evidence collection while it also performs external network access and persistence, violating expected trust boundaries and potentially leaking metadata or introducing supply-chain risk.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The skill is presented primarily as local Linux hardware inspection and fit validation, but the document also includes workflows that fetch replay inputs from GitHub and maintain a local cache. This behavior mismatch is dangerous because operators may approve the skill for local-only evidence collection while it also performs external network access and persistence, violating expected trust boundaries and potentially leaking metadata or introducing supply-chain risk.

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

The skill is presented primarily as local Linux hardware inspection and fit validation, but the document also includes workflows that fetch replay inputs from GitHub and maintain a local cache. This behavior mismatch is dangerous because operators may approve the skill for local-only evidence collection while it also performs external network access and persistence, violating expected trust boundaries and potentially leaking metadata or introducing supply-chain risk.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding

The skill requests or documents capabilities equivalent to shell execution, filesystem reads/writes, and network access, but it does not declare a constrained tool scope. That creates an unnecessary trust gap: a caller may invoke the skill expecting only local host inspection while the documented workflow also permits artifact creation and external fetching, increasing the chance of unintended data exposure or unsafe execution paths.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · SKILL.md (reported line 70)May include surrounding context.

Common setup (both routes, in a workspace outside the skill directory):

bash
mkdir -p reports

CLI (0.8.0): after install, export its bundled configuration once, in a

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The manifest describes fitctl as inspecting Linux hardware, collecting sensor evidence, monitoring observation windows, validating workload requirements, and comparing hosts using JSON. This file instead implements remote retrieval of example inputs via urllib from raw.githubusercontent.com, which is a separate content-fetching capability rather than an obvious part of local host inspection behavior.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · references/installation.md (reported line 53)May include surrounding context.

bash
FITCTL_INSTALL_ROOT="$PWD/.tools/fitctl-0.8.0"     # isolated root
BUILD_DIR=$(mktemp -d "${TMPDIR:-/tmp}/fitctl-install.XXXXXX")   # operation-owned scratch
mkdir -p "$BUILD_DIR/target"
findmnt -T "$BUILD_DIR"    # inspect the filesystem backing the build directory
CARGO_HOME="$PWD/.tools/cargo-home" CARGO_TARGET_DIR="$BUILD_DIR/target" \
  cargo install fitctl --version 0.8.0 --locked \

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/observe.py (reported line 76)May include surrounding context.

python
entry = {"index": index, "started_at": utc_now()}
        proc = None
        try:
            proc = subprocess.Popen([sys.executable, "-B", "-c", worker, json.dumps(options)],
                                    stdout=subprocess.PIPE, stderr=subprocess.PIPE,
                                    text=True, start_new_session=True)
            stdout, stderr = proc.communicate(timeout=min(args.timeout, remaining))

Scope Creep

Low
Category
Excessive Agency
Confidence
70% confidence
Finding

Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.

Content

Scanner excerpt · LICENSE (reported line 12)May include surrounding context.

text
permit persons to whom the Software is furnished to do so.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED,
INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A
PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT
HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE