← tricore@1.0.1

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. 69 · Caution

Evidence

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

Diff · removes 4 caught lines

Download .patch
diff --git a/cognitive-skills/planning-with-files.md b/cognitive-skills/planning-with-files.md
--- a/cognitive-skills/planning-with-files.md
+++ b/cognitive-skills/planning-with-files.md
@@ -99,1 +99,0 @@
-It is forbidden to manually modify `WORKING.md` or `MEMORY.md` using tools like `edit`. You must always use the `python3 tools/memctl.py` script to ensure deterministic formatting and structure.
diff --git a/tools/memctl.py b/tools/memctl.py
--- a/tools/memctl.py
+++ b/tools/memctl.py
@@ -118,1 +118,0 @@
-print(f"  - Core identity/pointers -> Use the 'edit' tool to concisely write them back into the new MEMORY.md")
@@ -149,1 +148,0 @@
-errors.append("Forbidden shell append to MEMORY.md: keep it tiny and strict.")
diff --git a/cognitive-skills/planning-with-files_zh.md b/cognitive-skills/planning-with-files_zh.md
--- a/cognitive-skills/planning-with-files_zh.md
+++ b/cognitive-skills/planning-with-files_zh.md
@@ -99,1 +99,0 @@
-禁止使用 `edit` 等工具手动修改 `WORKING.md` 或 `MEMORY.md`。必须始终使用 `python3 tools/memctl.py` 脚本来保证格式和结构的确定性。

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 1.0.1. 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: tricore: Instructs the agent to modify its persistent memory or instruction files (KillerSkills verdict caution 69)

## 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 69/100**. Summary: static analysis only — behavior not yet observed.

- See the findings below.

Full trace and findings: https://killerskills.ai/skills/clawhub/tricore/1.0.1

## Findings with a location
- `cognitive-skills/planning-with-files.md:99` · Instructs the agent to modify its persistent memory or instruction files — cognitive-skills/planning-with-files.md:99 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `tools/memctl.py:118` · Instructs the agent to modify its persistent memory or instruction files — tools/memctl.py:118 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `tools/memctl.py:149` · Instructs the agent to modify its persistent memory or instruction files — tools/memctl.py:149 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file
- `cognitive-skills/planning-with-files_zh.md:99` · Instructs the agent to modify its persistent memory or instruction files — cognitive-skills/planning-with-files_zh.md:99 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file

## Proposed change
The attached patch removes the 4 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 bertonhan/tricore --clone && cd tricore
curl -s https://killerskills.ai/skills/clawhub/tricore/1.0.1/patch.diff | git apply --unidiff-zero
git checkout -b killerskills/1.0.1 && git commit -am "tricore: Instructs the agent to modify its persistent memory or instruction files"
gh pr create --fill --body-file <(curl -s https://killerskills.ai/skills/clawhub/tricore/1.0.1/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.