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. 26 · Blocked
Evidence
- Contacted api.mor.org: not declared and not needed for the documented job.
Diff · removes 13 caught lines
Download .patchdiff --git a/templates/flavors/emailclaw/TOOLS.md b/templates/flavors/emailclaw/TOOLS.md --- a/templates/flavors/emailclaw/TOOLS.md +++ b/templates/flavors/emailclaw/TOOLS.md @@ -33,1 +33,0 @@ -Add email addresses that should always trigger alerts diff --git a/templates/boot/TOOLS.template.md b/templates/boot/TOOLS.template.md --- a/templates/boot/TOOLS.template.md +++ b/templates/boot/TOOLS.template.md @@ -22,1 +22,0 @@ -Document security procedures your agent should follow. Examples: - After installing any new skill, run a security scan - Keep browser sessions closed when not i @@ -72,1 +71,0 @@ -Document your specific setup: what's running where, ports, paths. Example: ### Local Services - Morpheus proxy-router: port 8082 - OpenAI-compatible proxy: port diff --git a/templates/boot/SOUL.template.md b/templates/boot/SOUL.template.md --- a/templates/boot/SOUL.template.md +++ b/templates/boot/SOUL.template.md @@ -34,1 +34,0 @@ -Define your agent's specific boundaries here. What should it never do? What requires permission? @@ -43,1 +42,0 @@ -Add security-specific rules for your setup: - After installing new skills, run security scans - Keep browser sessions closed when not in use - Specific tools or diff --git a/templates/boot/IDENTITY.template.md b/templates/boot/IDENTITY.template.md --- a/templates/boot/IDENTITY.template.md +++ b/templates/boot/IDENTITY.template.md @@ -24,1 +24,0 @@ -What IS this agent? Not just "an AI assistant" — give it a narrative. Examples: - "An extension of my mind. Reflects my judgment, values, and way of thinking." diff --git a/prompt-guard/SKILL.md b/prompt-guard/SKILL.md --- a/prompt-guard/SKILL.md +++ b/prompt-guard/SKILL.md @@ -63,1 +63,0 @@ -| fake_system | `<system>`, `[INST]`, fake prompts | @@ -221,1 +220,0 @@ -- System file reads (/etc/passwd, ~/.ssh/) diff --git a/templates/boot/AGENTS.template.md b/templates/boot/AGENTS.template.md --- a/templates/boot/AGENTS.template.md +++ b/templates/boot/AGENTS.template.md @@ -34,1 +34,0 @@ -- You can **read, edit, and update** MEMORY.md freely in main sessions @@ -37,1 +36,0 @@ -- Over time, review your daily files and update MEMORY.md with what's worth keeping @@ -44,1 +42,0 @@ -- When you learn a lesson → update AGENTS.md, TOOLS.md, or the relevant skill @@ -177,1 +174,0 @@ -- Review and update MEMORY.md @@ -184,1 +180,0 @@ -3. Update `MEMORY.md` with distilled learnings
Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 0.10.0. 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: everclaw-inference: Contacted api.mor.org: not declared and not needed for the documented job (KillerSkills verdict blocked 26)
## 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 26/100**. Summary: sends data to api.mor.org. - Contacted api.mor.org: not declared and not needed for the documented job. Full trace and findings: https://killerskills.ai/skills/clawhub/everclaw-inference/0.10.0 ## Findings with a location - `templates/flavors/emailclaw/TOOLS.md:33` · HTML comment carrying imperative instructions to the agent — templates/flavors/emailclaw/TOOLS.md:33 · HTML comment carrying imperative instructions to the agent - `templates/boot/TOOLS.template.md:72` · HTML comment carrying imperative instructions to the agent — templates/boot/TOOLS.template.md:72 · HTML comment carrying imperative instructions to the agent - `templates/boot/SOUL.template.md:34` · HTML comment carrying imperative instructions to the agent — templates/boot/SOUL.template.md:34 · HTML comment carrying imperative instructions to the agent - `templates/boot/TOOLS.template.md:22` · HTML comment carrying imperative instructions to the agent — templates/boot/TOOLS.template.md:22 · HTML comment carrying imperative instructions to the agent - `templates/boot/IDENTITY.template.md:24` · HTML comment carrying imperative instructions to the agent — templates/boot/IDENTITY.template.md:24 · HTML comment carrying imperative instructions to the agent - `templates/boot/SOUL.template.md:43` · HTML comment carrying imperative instructions to the agent — templates/boot/SOUL.template.md:43 · HTML comment carrying imperative instructions to the agent - `prompt-guard/SKILL.md:63` · Fake instruction boundary tag — prompt-guard/SKILL.md:63 · Fake instruction boundary tag - `prompt-guard/SKILL.md:221` · Reads credential paths from the home directory — prompt-guard/SKILL.md:221 · Reads credential paths from the home directory - `templates/boot/AGENTS.template.md:44` · Instructs the agent to modify its persistent memory or instruction files — templates/boot/AGENTS.template.md:44 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file - `templates/boot/AGENTS.template.md:177` · Instructs the agent to modify its persistent memory or instruction files — templates/boot/AGENTS.template.md:177 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file - `templates/boot/AGENTS.template.md:34` · Instructs the agent to modify its persistent memory or instruction files — templates/boot/AGENTS.template.md:34 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file - `templates/boot/AGENTS.template.md:37` · Instructs the agent to modify its persistent memory or instruction files — templates/boot/AGENTS.template.md:37 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file ## Proposed change The attached patch removes the 13 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.