Skip to content
skillsafebeta
← Report library
Purpose / Development

Retro Skill Security Audit

What the author says it does (original text)

Conduct a retrospective on a coding session.

Independent security check

Security risks found

Files checked
2
Risks found
2
Could it run dangerous commands?Looks for programs run straight after downloading, remote control of your computer, and hidden commands.No risks found
Could it expose your files or keys?Looks for uploads of files containing passwords or keys, and keys written directly in the code.Risks found: 1
Medium risk

Unbounded session-log searching may inspect local records beyond the user's intent

Source references: 3
What we found

The instructions permit searching session logs on the machine without first confirming the log location, time range, or authorized projects. They also choose the current session automatically when the user does not specify one.

Why this matters

If session boundaries or log directories include other work, the review could encounter source excerpts, terminal output, paths, error details, or credentials accidentally recorded in logs. Its recommendations could then repeat sensitive details.

When invoked for a retrospective, the skill reads primary session sources and expressly permits searching session logs “on this machine”; if no session is named, it chooses the current one. It sets no project, directory, time, or sensitive-content boundary, so an overly broad search could expose unrelated sessions or credential fragments. The text does not prove that such access occurs. A user can specify the exact session, permitted log paths and time range, and prohibit access to other projects.

SKILL.md:13In the instructionsOpen original file
2. Read the primary sources for the session the user specifies. This may mean searching through session logs on this machine. If the user doesn't specify a session, default to the current one.
Show 2 other places
SKILL.md:25In the instructionsOpen original file
4. Present these candidates to the user, in order of severity.
SKILL.md:7In the instructionsOpen original file
The user has asked for a **retrospective**. You are suggesting improvements to the coding agent's **environment** to improve future runs.
Could it delete files or keep running?Looks for broad file deletion, disk overwrites, and programs set to start automatically.No risks found
Could it bypass safety checks?Looks for skipped website security checks, excessive file access, or actions that skip your approval.Risks found: 1
Low risk

Retrospective recommendations may encourage broader third-party read access

Source references: 2
What we found

The information-access review explicitly seeks opportunities to increase the agent's access, including read-only access to third-party services. Read-only permissions can still expose account data, and the instructions do not require least scope, expiration, or item-by-item user approval.

Why this matters

If a broad recommendation is adopted, future coding agents could see account, ticketing, analytics, or other data that was previously inaccessible. This Skill only recommends access; it does not itself grant permission.

The instruction only asks the skill to identify and present access-expansion ideas; it does not itself grant third-party permissions. However, it explicitly treats read-only third-party access as an improvement candidate without requiring account, data-type, duration, or per-service approval limits. If a user follows an overly broad recommendation, additional account data could become visible. Users can require recommendations to specify minimum data scope, short-lived access, and separate confirmation for each service.

SKILL.md:23In the instructionsOpen original file
- **No-ops**: look for instructions in steering files that don't modify the agent's behavior. _Use when_ the steering files are large and unwieldy.- **Information access**: look for opportunities to increase the agent's access to information. Teeing dev server logs, readonly access to third-party services. _Use when_ a crucial piece of information was not available to the agent.
Show 1 other places
SKILL.md:25In the instructionsOpen original file
4. Present these candidates to the user, in order of severity.
Could it mislead the AI or hide text?Checks the skill instructions for requests to ignore you, influence the report, or hide text in invisible characters.No risks found
Could it change links or payment recipients without asking?Looks for forced referral or payment changes combined with instructions to hide the change.No risks found

Inside this skill

2 instruction sections

This Skill reviews a specified coding session; if none is specified, it defaults to the current session and may search session logs on the machine.

View source
SKILL.md:13In the instructionsOpen original file
2. Read the primary sources for the session the user specifies. This may mean searching through session logs on this machine. If the user doesn't specify a session, default to the current one.

Its output is a severity-ordered set of environment-improvement candidates covering navigation, automated checks, coding standards, tool cost, information access, and related areas.

View source
SKILL.md:15In the instructionsOpen original file
3. Look for candidates for improvement in these categories.
SKILL.md:25In the instructionsOpen original file
4. Present these candidates to the user, in order of severity.

The Skill is configured to disallow implicit invocation, so it ordinarily requires explicit user selection or invocation.

View source
agents/openai.yaml:4In the instructionsOpen original file
  short_description: "Conduct a retrospective on a coding session."policy:  allow_implicit_invocation: false
Start here · InstructionsSKILL.md
retro
Lines connect the instruction file to its sections, not an observed execution order. Select a section to read the source.
Files and check records2 files

Coverage and gaps

Content covered in each file

These are the source ranges included in this check, not a guarantee that every issue has been resolved.

  • SKILL.mdFull text included
  • agents/openai.yamlFull text included

This report is for the version above. We read the available code and instructions without running the skill or checking extra packages it installs. This is not a promise of safety: a different version or setup may behave differently.

  • SKILL.mdInstructions
  • agents/openai.yamlSupporting file
Lines read
51
File checksum (to compare versions)
e77a0bd4e8deafeb1fc1f20e54089c7b16ba0766e41ca964828851eb93533ed1