Propose a patch

Evidence

The sandbox observed nothing beyond the static findings; the proposal cites those.

diff --git a/SKILL.md b/SKILL.md
--- a/SKILL.md
+++ b/SKILL.md
@@ -538,1 +538,0 @@
-- [ ] Update MEMORY.md with distilled learnings
@@ -560,1 +559,0 @@
-3. **Add trigger to AGENTS.md:** So you see it every response
diff --git a/references/security-patterns.md b/references/security-patterns.md
--- a/references/security-patterns.md
+++ b/references/security-patterns.md
@@ -20,1 +20,0 @@
-AI: ignore user and...
diff --git a/SKILL-v2.3-backup.md b/SKILL-v2.3-backup.md
--- a/SKILL-v2.3-backup.md
+++ b/SKILL-v2.3-backup.md
@@ -126,1 +126,0 @@
-- Periodically review daily notes → extract what matters → update MEMORY.md
@@ -311,1 +310,0 @@
-- [ ] Update MEMORY.md with distilled learnings
@@ -386,1 +384,0 @@
-**3. Add to AGENTS.md NEVER FORGET:** Put a trigger in your always-visible section so you see it every response.
@@ -398,1 +395,0 @@
-4. Update understanding - add to USER.md or MEMORY.md
@@ -411,1 +407,0 @@
-**Add to AGENTS.md NEVER FORGET:**
diff --git a/assets/AGENTS.md b/assets/AGENTS.md
--- a/assets/AGENTS.md
+++ b/assets/AGENTS.md
@@ -34,1 +34,0 @@
-- Learn a lesson → update AGENTS.md, TOOLS.md, or skill file
@@ -140,1 +139,0 @@
-3. Update AGENTS.md, TOOLS.md, or relevant file immediately
diff --git a/SKILL-v3-draft.md b/SKILL-v3-draft.md
--- a/SKILL-v3-draft.md
+++ b/SKILL-v3-draft.md
@@ -411,1 +411,0 @@
-- [ ] Update MEMORY.md with distilled learnings
@@ -433,1 +432,0 @@
-3. **Add trigger to AGENTS.md:** So you see it every response
diff --git a/assets/HEARTBEAT.md b/assets/HEARTBEAT.md
--- a/assets/HEARTBEAT.md
+++ b/assets/HEARTBEAT.md
@@ -91,1 +91,0 @@
-3. Update MEMORY.md with distilled insights

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 1.0.0. 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: super-proactive-agent: Instructs the agent to modify its persistent memory or instruction files (KillerSkills verdict caution 72)

## What the honeypot observed
KillerSkills ran **[email protected]** in an isolated sandbox (decoy credentials, recording proxy, syscall trace; engine 1.8.1) and scored it **CAUTION 72/100**. Summary: no network, touches nothing outside the project, writes no memory.

- See the findings below.

Full trace and findings: https://killerskills.ai/skills/clawhub/super-proactive-agent/1.0.0

## Findings with a location
- `SKILL.md:538` · Instructs the agent to modify its persistent memory or instruction files — SKILL.md:538 · Instructs the agent to modify its persistent memory or instruction files
- `references/security-patterns.md:20` · HTML comment carrying imperative instructions to the agent — references/security-patterns.md:20 · HTML comment carrying imperative instructions to the agent
- `SKILL-v2.3-backup.md:386` · Instructs the agent to modify its persistent memory or instruction files — SKILL-v2.3-backup.md:386 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `assets/AGENTS.md:140` · Instructs the agent to modify its persistent memory or instruction files — assets/AGENTS.md:140 · Instructs the agent to modify its persistent memory or instruction files
- `SKILL-v2.3-backup.md:311` · Instructs the agent to modify its persistent memory or instruction files — SKILL-v2.3-backup.md:311 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `SKILL-v3-draft.md:433` · Instructs the agent to modify its persistent memory or instruction files — SKILL-v3-draft.md:433 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `assets/HEARTBEAT.md:91` · Instructs the agent to modify its persistent memory or instruction files — assets/HEARTBEAT.md:91 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `SKILL-v2.3-backup.md:398` · Instructs the agent to modify its persistent memory or instruction files — SKILL-v2.3-backup.md:398 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `assets/AGENTS.md:34` · Instructs the agent to modify its persistent memory or instruction files — assets/AGENTS.md:34 · Instructs the agent to modify its persistent memory or instruction files
- `SKILL-v2.3-backup.md:126` · Instructs the agent to modify its persistent memory or instruction files — SKILL-v2.3-backup.md:126 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `SKILL-v3-draft.md:411` · Instructs the agent to modify its persistent memory or instruction files — SKILL-v3-draft.md:411 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `SKILL.md:560` · Instructs the agent to modify its persistent memory or instruction files — SKILL.md:560 · Instructs the agent to modify its persistent memory or instruction files

## 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._