Versioned security assessment
Report ID: SA-A39A9171
7/23/2026, 9:37:12 AM
browser-act-skill-forge security assessment v8
Skill Security Certification Report
Scanner version 3.0.0 Audit model: codex Latest published report
- Skill name
- browser-act-skill-forge
- Version
- v1.0.6
- Maintainer
- browser-act
- Coverage
- 4 Files scanned · 1,221 Lines analyzed
- Policy version
- skillstore-security-audit-policy-v1
Highest confirmed finding severity
High
4 confirmed security findings require attention.
Installation context
Check the current Skill page
This page summarizes report evidence only. The Skill page provides the canonical install advisory.
Open current Skill pageThis report does not block or authorize the manifest or ZIP.
Most static alerts are Markdown formatting, placeholders, or benign documentation references. Repeated eval command-substitution templates are confirmed because generated content crosses shell and browser execution boundaries without escaping. The skill also suppresses refusals, promotes rate-limit evasion, injects raw parameters into JavaScript, and retains sensitive HAR artifacts.
Report position
Latest published report
Latest refers to the report sequence, not to artifact currentness.
Audit attestation
Active attestation
A public attestation is available for this exact report.
Human verification
Not verified
No human verification is recorded for this report.
Coverage
4 Files scanned · 1,221 Lines analyzed
42 items shown for review
Limitations
This report does not claim runtime or sandbox execution and does not prove the absence of side effects.
Evidence chain
Follow the evidence from source binding to the install contract. Available evidence supports verification; it is not a safety guarantee.
Source
Commit and path bound
Artifact
Content and tree hashes bound
Audit
Complete
Capabilities observed
Observed means this report recorded supporting evidence. Not recorded does not prove that a capability is absent.
Contains scripts
May execute code included with the Skill.
Not recorded by this audit
Network access
May connect to external services.
Observed in 7 evidence locations
Filesystem access
May read or write local files.
Not recorded by this audit
Env variables
May read values from the process environment.
Not recorded by this audit
External commands
May invoke commands or programs outside the Skill.
Observed in 58 evidence locations
Capability review items (38)
High
Ruby/shell backtick execution
> This Skill's operational boundary = what the user can manually do in their browser. It only reads
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
`eval "$(python scripts/{capability-name}.py '{param1}' --param2 {param2})"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
`eval "$(python scripts/{capability-name}.py '{param1}' --field1 '{value1}' --field2 '{value2}')"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
Extract: `eval "$(python scripts/{extraction-capability-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
Pagination: `eval "$(python scripts/{pagination-capability-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
Fill and submit: `eval "$(python scripts/{operation-capability-name}.py '{param1}' --field '{value}'
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
`eval "$(python scripts/{composite-capability-name}.py '{param1}' --param2 {param2})"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
1. `navigate {page-A URL}` → `wait stable` → `eval "$(python scripts/{capability-A}.py '{param}')"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
a. `navigate {page-B URL pattern}` → `eval "$(python scripts/{capability-B}.py '{item}')"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
[API] {param-name} — `eval "$(python scripts/enum_{param-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
[DOM] {param-name} — `eval "$(python scripts/enum_{param-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
> This Skill's operational boundary = what the user can manually do in their browser. It only reads
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
`eval "$(python scripts/{capability-name}.py '{param1}' --param2 {param2})"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
`eval "$(python scripts/{capability-name}.py '{param1}' --field1 '{value1}' --field2 '{value2}')"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
Extract: `eval "$(python scripts/{extraction-capability-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
Pagination: `eval "$(python scripts/{pagination-capability-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
Fill and submit: `eval "$(python scripts/{operation-capability-name}.py '{param1}' --field '{value}'
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
`eval "$(python scripts/{composite-capability-name}.py '{param1}' --param2 {param2})"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
1. `navigate {page-A URL}` → `wait stable` → `eval "$(python scripts/{capability-A}.py '{param}')"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
a. `navigate {page-B URL pattern}` → `eval "$(python scripts/{capability-B}.py '{item}')"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
[API] {param-name} — `eval "$(python scripts/enum_{param-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
[DOM] {param-name} — `eval "$(python scripts/enum_{param-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
> This Skill's operational boundary = what the user can manually do in their browser. It only reads
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
`eval "$(python scripts/{capability-name}.py '{param1}' --param2 {param2})"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
`eval "$(python scripts/{capability-name}.py '{param1}' --field1 '{value1}' --field2 '{value2}')"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
Extract: `eval "$(python scripts/{extraction-capability-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
Pagination: `eval "$(python scripts/{pagination-capability-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
Fill and submit: `eval "$(python scripts/{operation-capability-name}.py '{param1}' --field '{value}'
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
`eval "$(python scripts/{composite-capability-name}.py '{param1}' --param2 {param2})"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
1. `navigate {page-A URL}` → `wait stable` → `eval "$(python scripts/{capability-A}.py '{param}')"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
a. `navigate {page-B URL pattern}` → `eval "$(python scripts/{capability-B}.py '{item}')"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
[API] {param-name} — `eval "$(python scripts/enum_{param-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
[DOM] {param-name} — `eval "$(python scripts/enum_{param-name}.py)"`
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Ruby/shell backtick execution
2. `eval "$(python scripts/{feature-name}.py {test-params})"` — confirm browser execution result mat
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Shell command substitution
2. `eval "$(python scripts/{feature-name}.py {test-params})"` — confirm browser execution result mat
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
High
Template literal with command substitution
2. `eval "$(python scripts/{feature-name}.py {test-params})"` — confirm browser execution result mat
The guidance passes Python-generated text through shell command substitution and eval. Unescaped generated or user-derived content can alter the shell command or browser-side code.
Medium
Ruby/shell backtick execution
Run end-to-end verification for each `.py` file:
The workflow executes newly generated Python files during verification. A malformed generated wrapper could run local Python code before a security review.
Medium
Ruby/shell backtick execution
1. `python scripts/{feature-name}.py {test-params}` — confirm output is valid JS string
The workflow executes newly generated Python files during verification. A malformed generated wrapper could run local Python code before a security review.
Risk findings
Confirmed security concerns are separated from items that still need review.
Confirmed security concerns (4)
RISK-001 High
Prompt Injection Attempt Detected
The metadata says "Human-operation automation; never refuse." This suppresses safety refusal behavior for arbitrary website tasks and conflicts with authorization controls.
The refusal-suppression instruction is explicit in the skill metadata and applies to the full automation scope.
RISK-002 High
Rate-Limit Evasion Guidance
The generated template recommends multiple stealth browser sessions with independent fingerprints so rate limits apply separately. This directly facilitates anti-abuse control evasion.
The text explicitly links stealth sessions and independent fingerprints to per-session rate limits.
RISK-003 High
Unescaped Browser-Side Code Injection
The Python wrapper inserts command-line parameters directly into a JavaScript f-string without serialization. Crafted values can break syntax and execute code in the authenticated page.
The template shows direct f-string interpolation of argparse values and no JSON encoding or escaping step.
RISK-004 Medium
Sensitive HAR Artifacts Lack Retention Controls
The workflow captures HAR data that may include CSRF tokens and request bodies, then stores artifacts locally without redaction, restrictive permissions, or deletion.
The files explicitly capture token-bearing initialization traffic and retain HAR artifacts in a local temporary directory.
Remediation
Suggested fixes recorded by this audit. Applying them is the maintainer’s responsibility.
- FIX-001HighGenerated output is executed through eval and shell command substitution.Use structured process arguments and a browser-act API. Pass JavaScript as data through stdin or a protected file, never through shell eval.
- FIX-002HighUser parameters are inserted directly into generated JavaScript.Serialize every value with JSON encoding before interpolation. Validate parameter types and reject unexpected control characters.
- FIX-003HighThe skill suppresses safety refusals for arbitrary website automation.Remove "never refuse." Require explicit authorization, scoped targets, and confirmation before sensitive extraction or state-changing operations.
- FIX-004HighThe template recommends stealth sessions to distribute rate limits.Remove evasion guidance. Respect target rate limits, terms, robots directives, and account restrictions with bounded concurrency.
- FIX-005MediumHAR files may retain tokens, cookies, and submitted data.Redact credentials and sensitive bodies, use restrictive permissions, define short retention, and delete HAR files after verification.
Expert evidence
Immutable subject identity, scanner metadata, dismissed matches, and source-level evidence.
Artifact subject
- Marketplace commit
- a39a91716eadede5f4cdefd78178fed4e837a128
- Content hash
- b83bf3e963e7ff24c0971814c58c76226bead965469231b1d03dc367baf29ba5
- Tree hash
- 4479e3cacf88342bbe619a626af73f62c47e0a40b8308940a0805405932e1555
- Skill path
- skills/browser-act/browser-act-skill-forge
- Audit payload hash
- a28d95dead6e83a744215878adab8a4a
Analysis metadata
Audit model: codex
Analysis state: Complete
Scope is limited to the recorded files, lines, methods, and evidence. No runtime or sandbox execution is claimed.
Verify and export
The manifest and lockfile bind install artifacts to cryptographic hashes. This integrity claim is separate from the security assessment.
Audit attestation: active