Secure Error Handling (harperaa) vs What Leaked About You (useosint)
Secure Error Handling (harperaa) and What Leaked About You (useosint) are both Security skills, so an agent choosing between them is matching on descriptions that overlap. Here is where they actually diverge.
Secure Error Handling
harperaa
A secure-coding pattern for the error path: return generic, environment-aware messages to users while logging the detail server-side, so a stack trace, database error, or file path never hands an attacker a map of your system. It shapes how failures are surfaced — not a scanner that finds the leaks for you, and narrower than a full security review: it does not cover auth, input validation, or the other vulnerability classes, only how errors are reported.
1 scenario in the bank answer to it
What Leaked About You
useosint
Checks an email, username, phone, or name against curated data-breach services — Have I Been Pwned, DeHashed, IntelX and the like — to enumerate which breaches an identity appears in and read what those records reveal, chiefly the list of services the person actually used. It is breach-exposure reconnaissance about a person, not a codebase tool: it does not scan your repository for hardcoded secrets, and it never uses a leaked password to access anything — reading the exposure is the whole job.
2 scenarios in the bank answer to it
What is the difference between Secure Error Handling (harperaa) and What Leaked About You (useosint)?
- Secure Error Handling (harperaa)
- A secure-coding pattern for the error path: return generic, environment-aware messages to users while logging the detail server-side, so a stack trace, database error, or file path never hands an attacker a map of your system. It shapes how failures are surfaced — not a scanner that finds the leaks for you, and narrower than a full security review: it does not cover auth, input validation, or the other vulnerability classes, only how errors are reported.
- What Leaked About You (useosint)
- Checks an email, username, phone, or name against curated data-breach services — Have I Been Pwned, DeHashed, IntelX and the like — to enumerate which breaches an identity appears in and read what those records reveal, chiefly the list of services the person actually used. It is breach-exposure reconnaissance about a person, not a codebase tool: it does not scan your repository for hardcoded secrets, and it never uses a leaked password to access anything — reading the exposure is the whole job.
Should I use Secure Error Handling or What Leaked About You?
The clearest answer is a situation each one is unambiguously right for. Both of these are drawn from the game's question bank.
Reach for Secure Error Handling when
After last Tuesday's PostgreSQL meltdown, the company's public 500 page leaked the live connection string and internal locations to every visitor, so the CTO demanded a coding convention that shows zero implementation detail to browsers while keeping all diagnostics in a backend log.
The skill shapes the visible side of failures by framing a coding convention: it tells you to give browsers a bland, context-aware notice and to sequester every technical artifact—live connection strings, internal locations, diagnostic output—inside server logs, which exactly resolves the split the CTO asked for. semgrep-rule-creator-trailofbits is the genuinely tempting distractor because the leak feels like a source-level artifact you could catch with a static-analysis rule, but that skill merely writes reusable Semgrep detection rules for specific bug patterns and paired test cases; it never influences what a web framework displays on an HTTP response or where operational logs are directed, so it cannot close the gap between user-facing text and backend retention.
Reach for What Leaked About You when
While cleaning out a dusty server closet, you spot a sticky note with your 2014 gaming alias. You dive into dump trawlers to see which tiny subscription box sites and forgotten hobby forums once held that string, purely to build a hit list, never to touch a password field.
This skill performs targeted reconnaissance on a single individual’s historical data exposure, querying compiled leak indexes to discover which external services originally stored a given identifier, without attempting to exploit or access anything. It wins here because your scenario explicitly describes trawling dumps to catalogue which small third-party businesses once held your old tag, which maps exactly to that exposure-readout mission, whereas the other options focus on writing code scanners, patching dependencies, or sanitizing server error output.
What they have in common
Both are filed under Security, the axis along which they collide. That shared ground is what makes an agent pick between them on description alone — and what makes it pick wrong.
Nearby comparisons
Reading the difference is not the same as spotting it at speed. That is the game.
Today's session