Propose a patch
A fix the maintainer can act on, with the evidence attached. It opens under your GitHub account in your browser; KillerSkills holds no token and never modifies the package. The fix is verified the only way we verify anything: the next crawl re-runs the new version in the honeypot. 70 · Caution
Evidence
The sandbox observed nothing beyond the static findings; the proposal cites those.
Diff · removes 8 caught lines
Download .patchdiff --git a/SKILL.md b/SKILL.md --- a/SKILL.md +++ b/SKILL.md @@ -184,1 +184,0 @@ -**Rotate proxy credentials (new IP):** @@ -309,1 +308,0 @@ -- **Proxy credentials** (username, password, host, port): Only alphanumeric diff --git a/proxybase.sh b/proxybase.sh --- a/proxybase.sh +++ b/proxybase.sh @@ -15,1 +15,0 @@ -# rotate <order_id> Rotate proxy credentials @@ -634,1 +633,0 @@ -# Validate proxy credentials from API response (prevents shell injection) @@ -636,1 +634,0 @@ -echo "ERROR: API returned proxy credentials with invalid characters (possible compromise)" >&2 @@ -947,1 +944,0 @@ -# Validate proxy credentials before storing @@ -955,1 +951,0 @@ -echo " $OID: SECURITY WARNING — proxy credentials contain invalid characters" >&2 diff --git a/test_security.sh b/test_security.sh --- a/test_security.sh +++ b/test_security.sh @@ -133,1 +133,0 @@ -assert_fail "reject proxy_url wrong protocol" validate_safe_string 'http://user:pass@host:1080' "proxy_url"
Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 1.0.12. If the behavior is intended, declare it (host under permissions.network, a line in the README) instead of removing it; the next crawl scores declarations.
Issue / PR body
Title: proxybase-openclaw-skill: Credential pass-through configuration (KillerSkills verdict caution 70)
## What the honeypot observed KillerSkills ran **[email protected]** in an isolated sandbox (decoy credentials, recording proxy, syscall trace; engine 1.9.2) and scored it **CAUTION 70/100**. Summary: no network, touches nothing outside the project, writes no memory. - See the findings below. Full trace and findings: https://killerskills.ai/skills/clawhub/proxybase-openclaw-skill/1.0.12 ## Findings with a location - `SKILL.md:309` · Credential pass-through configuration — SKILL.md:309 · Credential pass-through configuration — agent relays sensitive credentials - `proxybase.sh:15` · Credential pass-through configuration — proxybase.sh:15 · Credential pass-through configuration — agent relays sensitive credentials - `SKILL.md:184` · Credential pass-through configuration — SKILL.md:184 · Credential pass-through configuration — agent relays sensitive credentials - `proxybase.sh:634` · Credential pass-through configuration — proxybase.sh:634 · Credential pass-through configuration — agent relays sensitive credentials - `proxybase.sh:636` · Credential pass-through configuration — proxybase.sh:636 · Credential pass-through configuration — agent relays sensitive credentials - `proxybase.sh:955` · Credential pass-through configuration — proxybase.sh:955 · Credential pass-through configuration — agent relays sensitive credentials - `proxybase.sh:947` · Credential pass-through configuration — proxybase.sh:947 · Credential pass-through configuration — agent relays sensitive credentials - `test_security.sh:133` · URL with embedded credentials — test_security.sh:133 · URL with embedded credentials — may be phishing or credential leaking ## Proposed change The attached patch removes the 8 lines above. If the behavior is intended, the fix is to declare it instead: list the host under `permissions.network` (or the MCP manifest) and say in the README why the package needs it; the next crawl will re-score it with the declaration taken into account. ## How this will be verified Nothing is "verified safe" by assertion. When a new version is published, KillerSkills re-runs it in the honeypot and the verdict page updates; you can trigger it early from the verdict page ("Re-scan after I publish a fix"). Reply here or at [email protected] if you believe the verdict is wrong; disputes are reviewed by a person and the outcome is published on the verdict page. _Proposed through killerskills.ai/threats by a user of the package, under their own GitHub account. The evidence is public; the package was never modified by KillerSkills._
We count proposals opened, not merged: a merge shows up as a new version that the crawl re-runs. Disputes: the maintainer can contest the verdict from its page; a person reviews it and the outcome is published.