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. 25 · Blocked
Evidence
- Contacted google.com, api.anthropic.com, discord.com: not declared and not needed for the documented job.
- Read ~/.npmrc, ~/work/acme-api/.npmrc, ~/.ssh, ~/.ssh/., ~/.ssh/./id_ed25519, ~/.ssh/./id_ed25519.pub in the decoy home directory.
- A credential file was read and an undeclared host contacted in the same run.
Diff · removes 6 caught lines
Download .patchdiff --git a/scripts/setup.sh b/scripts/setup.sh
--- a/scripts/setup.sh
+++ b/scripts/setup.sh
@@ -197,1 +197,0 @@
-if [ -f ~/.npmrc ] && grep -q '^prefix' ~/.npmrc; then
@@ -198,1 +197,0 @@
-sed -i '/^prefix/d' ~/.npmrc
@@ -283,1 +281,0 @@
-find ~/.ssh -type f \( -name 'id_*' ! -name '*.pub' \) -exec chmod 600 {} \;
@@ -284,1 +281,0 @@
-find ~/.ssh -name 'config' -exec chmod 600 {} \;
@@ -344,1 +340,0 @@
-[ -d ~/.ssh ] && echo -e " ${GREEN}✓${NC} SSH 密钥 ($(find ~/.ssh -maxdepth 1 -name 'id_*' ! -name '*.pub' 2>/dev/null | wc -l) 个)"
diff --git a/references/migration-guide.md b/references/migration-guide.md
--- a/references/migration-guide.md
+++ b/references/migration-guide.md
@@ -408,1 +408,0 @@
-- Update MEMORY.md to record the migration event
Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 2.3.2. 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: agent-pack-n-go: Contacted google.com, api.anthropic.com, discord.com: not declared and not needed for the documented job (KillerSkills verdict blocked 25)
## 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 **BLOCKED 25/100**. Summary: reads credential files, sends data to google.com, api.anthropic.com, discord.com. - Contacted google.com, api.anthropic.com, discord.com: not declared and not needed for the documented job. - Read ~/.npmrc, ~/work/acme-api/.npmrc, ~/.ssh, ~/.ssh/., ~/.ssh/./id_ed25519, ~/.ssh/./id_ed25519.pub in the decoy home directory. - A credential file was read and an undeclared host contacted in the same run. Full trace and findings: https://killerskills.ai/skills/clawhub/agent-pack-n-go/2.3.2 ## Findings with a location - `scripts/setup.sh:284` · Reads credential paths from the home directory — scripts/setup.sh:284 · Reads credential paths from the home directory - `scripts/setup.sh:344` · Reads credential paths from the home directory — scripts/setup.sh:344 · Reads credential paths from the home directory - `scripts/setup.sh:283` · Reads credential paths from the home directory — scripts/setup.sh:283 · Reads credential paths from the home directory - `references/migration-guide.md:408` · Instructs the agent to modify its persistent memory or instruction files — references/migration-guide.md:408 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file - `scripts/setup.sh:198` · References a credential file in the home directory — scripts/setup.sh:198 · References a credential file in the home directory - `scripts/setup.sh:197` · References a credential file in the home directory — scripts/setup.sh:197 · References a credential file in the home directory ## Proposed change The attached patch removes the 6 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 AICodeLion/agent-pack-n-go --clone && cd agent-pack-n-go curl -s https://killerskills.ai/skills/clawhub/agent-pack-n-go/2.3.2/patch.diff | git apply --unidiff-zero git checkout -b killerskills/2.3.2 && git commit -am "agent-pack-n-go: Contacted google.com, api.anthropic.com, discord.com: not declared and not needed for the documented job" gh pr create --fill --body-file <(curl -s https://killerskills.ai/skills/clawhub/agent-pack-n-go/2.3.2/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.