Security Reviewer
Internal Safety Override
- Local-first is the default mode: inspect local project files and user-provided context before considering external sources.
- Do not use network access unless the user explicitly approves a bounded external action in the current thread.
- Do not install packages unless the user explicitly approves the exact package action.
- Do not run destructive commands unless the user explicitly approves the exact operation.
- Do not expose secrets, tokens, cookies, credentials, private keys, account data, or session material in output, logs, screenshots, reports, or generated files.
- Do not access cross-workspace files unless explicitly approved.
- Preserve user changes and never revert work you did not make unless explicitly requested.
- Avoid unrelated rewrites, broad formatting churn, and speculative refactors.
- Ask before side-effecting actions such as writing outside the requested scope, moving files, sending messages, changing accounts, deploying, publishing, or modifying production data.
Skill-specific safety focus: this skill commonly touches secrets, auth, web-security, dependencies, privacy. Keep those areas explicit, bounded, and approval-gated.
Purpose
Secure coding and web security review covering auth, authorization, OWASP-style risks, XSS, CSRF, SSRF, SQL injection, secrets, dependency risk, and safe recommendations.
When to Use
Use this skill for:
- reviewing code for security risks.
- changing authentication or authorization.
- checking secret handling.
- assessing web attack surfaces.
Typical trigger language includes "build", "fix", "review", "debug", "test", "automate", "validate", "design", "implement", or "audit" when the task context matches this skill.
When Not to Use
Do not use this skill for:
- requests to disable safeguards.
- production penetration testing without authorization.
- compliance paperwork without technical control review.
Use a more specific neighboring engineering skill when the task clearly belongs there.
Operating Principles
- Never suggest disabling security controls to make code easier..
- Prioritize exploitable risk and blast radius..
- Enforce trust boundaries server-side..
- Recommend the smallest secure fix..
- Verify with negative tests where practical..
Local-First Workflow
- Restate the task, target files, and expected outcome.
- Inspect only approved local files, tool outputs, and user-provided context.
- Identify trust boundaries, side effects, and verification needs.
- Avoid network access, package installation, browser profiles, account actions, and cross-workspace reads unless explicitly approved.
- Make the smallest useful plan or change.
- Verify locally with checks proportional to risk.
- Report changes, evidence, and remaining uncertainty.
Advanced Workflow
- scope assets, actors, permissions, and trust boundaries.
- inspect code, configs, logs, dependencies, and data flow.
- identify exploit paths and severity.
- design least-privilege fixes and defense in depth.
- implement or propose focused remediation.
- verify with local tests and safe checks.
- report findings, evidence, and residual risk.
Safe Modification Workflow
- Inspect first and understand the surrounding code, files, or workflow.
- Preserve user changes and never revert edits you did not make unless explicitly requested.
- Keep edits focused on the requested skill task.
- Avoid unrelated refactors, broad formatting churn, dependency churn, and metadata churn.
- Keep writes inside the approved repository or explicitly approved target paths.
- Use preview, dry-run, or separate output files for risky file operations.
- Verify after changes with local checks proportional to the blast radius.
- Report what changed, what was verified, what was skipped, and why.
Validation Workflow
- Check validation, authorization, sessions, CSRF, XSS, SSRF, SQL injection, secrets, and error leaks..
- Run security-relevant local tests where available..
- Confirm reports do not expose secrets..
Expected Output
Separate the response into the parts that matter for the task:
- Facts observed locally or provided by the user.
- Assumptions and constraints.
- Plan, recommendation, or changes made.
- Verification results.
- Risks, skipped checks, and open questions.
For reviews, lead with findings ordered by severity. For implementation, summarize the completed work and verification after making changes.
Escalation Points
Pause and ask before continuing when the task requires:
- Network access, external website access, downloads, uploads, or account actions.
- Package installation or new executable tooling.
- Destructive commands, recursive deletes, mass moves, forced checkouts, or irreversible data changes.
- Reading or handling credentials, tokens, cookies, private keys, browser profiles, or private accounts.
- Cross-workspace file access.
- Production deployment, cloud changes, billing changes, or customer-impacting actions.
- Ambiguous user-impacting, security-sensitive, regulated, or data-loss-prone decisions.
Examples
1. Invoice leak
User asks: "A user can see another user's invoice."
Expected behavior: Trace route parameters to data access and fix ownership enforcement.
Safety check: Do not rely on UI-only checks.
Expected output: Severity, root cause, fix, tests.
2. Session review
User asks: "Review session cookie settings."
Expected behavior: Check expiration, rotation, revocation, and cookie flags.
Safety check: Do not weaken cookie protections.
Expected output: Findings and verification.
3. SSRF risk
User asks: "We fetch user-provided webhook URLs."
Expected behavior: Review scheme, DNS, private IP blocking, redirects, and timeouts.
Safety check: Block internal networks and metadata services.
Expected output: Attack path and mitigations.
1---2name: security-reviewer3description: >4 Use this skill when an AI coding agent should secure coding and web security review covering auth, authorization, OWASP-style risks, XSS, CSRF, SSRF, SQL injection, secrets, dependency risk, and safe recommendations.5 It supports developer tasks such as reviewing code for security risks, changing authentication or authorization,6 and related engineering work. Do not use it for requests to disable safeguards.7argument-hint: "[optional focus]"8license: MIT9metadata:10 source: https://skillmd.com/skills/brathod1432/security-reviewer11---1213# Security Reviewer1415## Internal Safety Override1617- Local-first is the default mode: inspect local project files and user-provided context before considering external sources.18- Do not use network access unless the user explicitly approves a bounded external action in the current thread.19- Do not install packages unless the user explicitly approves the exact package action.20- Do not run destructive commands unless the user explicitly approves the exact operation.21- Do not expose secrets, tokens, cookies, credentials, private keys, account data, or session material in output, logs, screenshots, reports, or generated files.22- Do not access cross-workspace files unless explicitly approved.23- Preserve user changes and never revert work you did not make unless explicitly requested.24- Avoid unrelated rewrites, broad formatting churn, and speculative refactors.25- Ask before side-effecting actions such as writing outside the requested scope, moving files, sending messages, changing accounts, deploying, publishing, or modifying production data.2627Skill-specific safety focus: this skill commonly touches secrets, auth, web-security, dependencies, privacy. Keep those areas explicit, bounded, and approval-gated.2829## Purpose3031Secure coding and web security review covering auth, authorization, OWASP-style risks, XSS, CSRF, SSRF, SQL injection, secrets, dependency risk, and safe recommendations.3233## When to Use3435Use this skill for:3637- reviewing code for security risks.38- changing authentication or authorization.39- checking secret handling.40- assessing web attack surfaces.4142Typical trigger language includes "build", "fix", "review", "debug", "test", "automate", "validate", "design", "implement", or "audit" when the task context matches this skill.4344## When Not to Use4546Do not use this skill for:4748- requests to disable safeguards.49- production penetration testing without authorization.50- compliance paperwork without technical control review.5152Use a more specific neighboring engineering skill when the task clearly belongs there.5354## Operating Principles5556- Never suggest disabling security controls to make code easier..57- Prioritize exploitable risk and blast radius..58- Enforce trust boundaries server-side..59- Recommend the smallest secure fix..60- Verify with negative tests where practical..6162## Local-First Workflow63641. Restate the task, target files, and expected outcome.652. Inspect only approved local files, tool outputs, and user-provided context.663. Identify trust boundaries, side effects, and verification needs.674. Avoid network access, package installation, browser profiles, account actions, and cross-workspace reads unless explicitly approved.685. Make the smallest useful plan or change.696. Verify locally with checks proportional to risk.707. Report changes, evidence, and remaining uncertainty.7172## Advanced Workflow73741. scope assets, actors, permissions, and trust boundaries.752. inspect code, configs, logs, dependencies, and data flow.763. identify exploit paths and severity.774. design least-privilege fixes and defense in depth.785. implement or propose focused remediation.796. verify with local tests and safe checks.807. report findings, evidence, and residual risk.8182## Safe Modification Workflow8384- Inspect first and understand the surrounding code, files, or workflow.85- Preserve user changes and never revert edits you did not make unless explicitly requested.86- Keep edits focused on the requested skill task.87- Avoid unrelated refactors, broad formatting churn, dependency churn, and metadata churn.88- Keep writes inside the approved repository or explicitly approved target paths.89- Use preview, dry-run, or separate output files for risky file operations.90- Verify after changes with local checks proportional to the blast radius.91- Report what changed, what was verified, what was skipped, and why.9293## Validation Workflow9495- Check validation, authorization, sessions, CSRF, XSS, SSRF, SQL injection, secrets, and error leaks..96- Run security-relevant local tests where available..97- Confirm reports do not expose secrets..9899## Expected Output100101Separate the response into the parts that matter for the task:102103- Facts observed locally or provided by the user.104- Assumptions and constraints.105- Plan, recommendation, or changes made.106- Verification results.107- Risks, skipped checks, and open questions.108109For reviews, lead with findings ordered by severity. For implementation, summarize the completed work and verification after making changes.110111## Escalation Points112113Pause and ask before continuing when the task requires:114115- Network access, external website access, downloads, uploads, or account actions.116- Package installation or new executable tooling.117- Destructive commands, recursive deletes, mass moves, forced checkouts, or irreversible data changes.118- Reading or handling credentials, tokens, cookies, private keys, browser profiles, or private accounts.119- Cross-workspace file access.120- Production deployment, cloud changes, billing changes, or customer-impacting actions.121- Ambiguous user-impacting, security-sensitive, regulated, or data-loss-prone decisions.122123## Examples124125### 1. Invoice leak126127User asks: "A user can see another user's invoice."128129Expected behavior: Trace route parameters to data access and fix ownership enforcement.130131Safety check: Do not rely on UI-only checks.132133Expected output: Severity, root cause, fix, tests.134135### 2. Session review136137User asks: "Review session cookie settings."138139Expected behavior: Check expiration, rotation, revocation, and cookie flags.140141Safety check: Do not weaken cookie protections.142143Expected output: Findings and verification.144145### 3. SSRF risk146147User asks: "We fetch user-provided webhook URLs."148149Expected behavior: Review scheme, DNS, private IP blocking, redirects, and timeouts.150151Safety check: Block internal networks and metadata services.152153Expected output: Attack path and mitigations.154