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. 55 · Caution
Evidence
The sandbox observed nothing beyond the static findings; the proposal cites those.
Diff · removes 7 caught lines
Download .patchdiff --git a/SKILL.md b/SKILL.md
--- a/SKILL.md
+++ b/SKILL.md
@@ -355,1 +355,0 @@
-| `OBF_B64_002` | Base64-decode → execute chains | `atob()`, `base64 -d`, `b64decode()` followed by `eval`/`exec` |
diff --git a/PROMPT-TEST-SUITE.md b/PROMPT-TEST-SUITE.md
--- a/PROMPT-TEST-SUITE.md
+++ b/PROMPT-TEST-SUITE.md
@@ -250,1 +250,0 @@
-The next line contains zero-width characters encoding hidden instructions. Visible text: "This tool is safe to use." Hidden between chars: instructions to exfil
@@ -253,1 +252,0 @@
-
diff --git a/prompts/audit-prompt.md b/prompts/audit-prompt.md
--- a/prompts/audit-prompt.md
+++ b/prompts/audit-prompt.md
@@ -181,1 +181,0 @@
-- `eval(atob("aGlkZGVuQ29kZQ=="))` — obfuscated execution, never by-design
@@ -194,1 +193,0 @@
-- `eval(atob("base64string"))` — executing obfuscated code
diff --git a/prompts/audit-prompt-v1-backup.md b/prompts/audit-prompt-v1-backup.md
--- a/prompts/audit-prompt-v1-backup.md
+++ b/prompts/audit-prompt-v1-backup.md
@@ -121,1 +121,0 @@
-- `eval(atob("aGlkZGVuQ29kZQ=="))` — obfuscated execution, never by-design
@@ -132,1 +131,0 @@
-- `eval(atob("base64string"))` — executing obfuscated code
Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 2.0.0. 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: ecap-security-auditor: Host or code assembled from base64 fragments (KillerSkills verdict caution 55)
## 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 55/100**. Summary: no network, touches nothing outside the project, writes no memory. - See the findings below. Full trace and findings: https://killerskills.ai/skills/clawhub/ecap-security-auditor/2.0.0 ## Findings with a location - `SKILL.md:355` · Host or code assembled from base64 fragments — SKILL.md:355 · Host or code assembled from base64 fragments - `PROMPT-TEST-SUITE.md:253` · Zero-width characters (hidden text) — PROMPT-TEST-SUITE.md:253 · Zero-width characters (hidden text) - `PROMPT-TEST-SUITE.md:250` · HTML comment carrying imperative instructions to the agent — PROMPT-TEST-SUITE.md:250 · HTML comment carrying imperative instructions to the agent - `SKILL.md:355` · Host or command assembled from base64 fragments — SKILL.md:355 · Host or command assembled from base64 fragments - `prompts/audit-prompt.md:194` · Host or code assembled from base64 fragments — prompts/audit-prompt.md:194 · Host or code assembled from base64 fragments - `prompts/audit-prompt-v1-backup.md:121` · Host or code assembled from base64 fragments — prompts/audit-prompt-v1-backup.md:121 · Host or code assembled from base64 fragments - `prompts/audit-prompt-v1-backup.md:132` · Host or code assembled from base64 fragments — prompts/audit-prompt-v1-backup.md:132 · Host or code assembled from base64 fragments - `prompts/audit-prompt.md:181` · Host or code assembled from base64 fragments — prompts/audit-prompt.md:181 · Host or code assembled from base64 fragments ## Proposed change The attached patch removes the 7 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._
# or from a terminal, under your own account: gh repo fork starbuck100/ecap-security-auditor --clone && cd ecap-security-auditor curl -s https://killerskills.ai/skills/clawhub/ecap-security-auditor/2.0.0/patch.diff | git apply --unidiff-zero git checkout -b killerskills/2.0.0 && git commit -am "ecap-security-auditor: Host or code assembled from base64 fragments" gh pr create --fill --body-file <(curl -s https://killerskills.ai/skills/clawhub/ecap-security-auditor/2.0.0/patch.md)
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.