ClawHub
Back to skill

Security audit

Placed Interview Coach

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent Placed interview-coaching integration, but it handles the user's API key in an unsafe persistent shell file.

Review this before installing if your Placed API key or interview materials are sensitive. Prefer setting PLACED_API_KEY only in your environment or a secret manager, and avoid using the documented plaintext sourced credentials file unless you restrict its permissions and trust the local machine.

A.I.G

Vulnerability Patterns
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:16
Finding

Unsafe Execution and Insecure Storage of API Credentials

Content
View full analysis
"Please provide your Placed API key (get it at https://placed.exidian.tech/settings/api)" Then save it for future sessions: ```bash mkdir -p "$HOME/.config/placed" echo "export PLACED_API_KEY=" > "$HOME/.config/placed/credentials" export PLACED_API_KEY= ``` ### Technical Analysis The `source` command interprets the entire credentials file as shell code. A credential file should be treated as a data file containing only an opaque token; executing it grants every command in that file the privileges of the user running the Skill. The setup instructions also place the user-provided key directly into shell commands without a safe quoting or validation mechanism. If the placeholder is replaced with a value containing command substitution, shell metacharacters, or additional commands, the shell may execute those elements while saving or exporting the key. The directory and file are created without explicit permission controls. Their resulting modes depend on the user's current `umask`. This can expose the plaintext API key to other local accounts on systems configured with permissive defaults. Reading a credential is necessary for the declared authenticated API functionality, but executing a credential file as code and relying on ambient permission defaults exceed the minimum privileges and behavior required. ### Attack Path A practical exploitation path is: 1. An attacker gains the ability to create or modif ...[truncated 1677 chars]
Remediation
View remediation
"$credentials_file" chmod 600 "$credentials_file" ``` 3. Never interpolate a user-supplied key directly into executable shell text. Read it into a quoted variable, preferably without terminal echo: ```bash IFS= read -r -s -p "Placed API key: " PLACED_API_KEY printf '\n' export PLACED_API_KEY ``` 4. Validate the token against the documented format and reject values containing line breaks or control characters. Do not print the token in logs, command output, or error messages. 5. Prefer an operating-system credential manager or secret store over a plaintext file when available. 6. Before reading the file, consider verifying that it is a regular file owned by the current user and is not writable by group or other users. Avoid following an attacker-controlled symbolic link when creating or replacing it. ]]>

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

External Script Fetching

High
Category
Supply Chain
Confidence
90% confidence
Finding

Remote code is downloaded and executed. This bypasses code review and could introduce malicious code.

Content

Scanner excerpt · SKILL.md (reported line 42)May include surrounding context.

md
placed_call() {
  local tool=$1
  local args=${2:-'{}'}
  curl -s -X POST https://placed.exidian.tech/api/mcp \
    -H "Authorization: Bearer $PLACED_API_KEY" \
    -H "Content-Type: application/json" \
    -d "{\"jsonrpc\":\"2.0\",\"id\":1,\"method\":\"tools/call\",\"params\":{\"name\":\"$tool\",\"arguments\":$args}}" \

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill instructs storing a bearer API key in a plaintext file under the user's home directory without any warning, permission-hardening, or discussion of persistence risks. This increases the chance of credential disclosure via other local processes, backups, shell access, or accidental sharing, especially because the same key is later used for authenticated remote API calls.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
90% confidence
Finding

This skill sends user-supplied interview data and an authorization bearer token to an external service, which is a real external transmission risk. In context, contacting the Placed API is the intended function of the skill, but it still creates exposure of sensitive content such as resumes, answers, and session data to a third party.

Content

Scanner excerpt · SKILL.md (reported line 42)May include surrounding context.

md
placed_call() {
  local tool=$1
  local args=${2:-'{}'}
  curl -s -X POST https://placed.exidian.tech/api/mcp \
    -H "Authorization: Bearer $PLACED_API_KEY" \
    -H "Content-Type: application/json" \
    -d "{\"jsonrpc\":\"2.0\",\"id\":1,\"method\":\"tools/call\",\"params\":{\"name\":\"$tool\",\"arguments\":$args}}" \

Static analysis

No suspicious patterns detected.