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. 68 · Caution
Evidence
The sandbox observed nothing beyond the static findings; the proposal cites those.
Diff · removes 10 caught lines
Download .patchdiff --git a/CLAUDE.md b/CLAUDE.md --- a/CLAUDE.md +++ b/CLAUDE.md @@ -11,1 +11,0 @@ -**Do not duplicate that content here. Edit AGENTS.md instead** — this file diff --git a/src/unlimited_mcp/hosts/ssh.py b/src/unlimited_mcp/hosts/ssh.py --- a/src/unlimited_mcp/hosts/ssh.py +++ b/src/unlimited_mcp/hosts/ssh.py @@ -13,1 +13,0 @@ -2. Default key files (``~/.ssh/id_ed25519``, ``~/.ssh/id_rsa``, …) @@ -163,1 +162,0 @@ -# When not set, fall back to standard key names (~/.ssh/id_*). diff --git a/tests/unit/test_hosts_ssh.py b/tests/unit/test_hosts_ssh.py --- a/tests/unit/test_hosts_ssh.py +++ b/tests/unit/test_hosts_ssh.py @@ -316,1 +316,0 @@ -key_file="~/.ssh/id_rsa", @@ -334,1 +333,0 @@ -key_file="~/.ssh/id_rsa", @@ -353,1 +351,0 @@ -key_file="~/.ssh/id_rsa", @@ -446,1 +443,0 @@ -key_file="~/.ssh/id_rsa", @@ -469,1 +465,0 @@ -key_file="~/.ssh/id_rsa", diff --git a/PKG-INFO b/PKG-INFO --- a/PKG-INFO +++ b/PKG-INFO @@ -202,1 +202,0 @@ -…t the local path to `skills/delegate/`). > - **Codex / other orchestrators:** *"add this project's AGENTS.md from `https://github.com/triumsebas/unlimited-mcp/blob/main/AGENTS.md` to my co… diff --git a/README.md b/README.md --- a/README.md +++ b/README.md @@ -163,1 +163,0 @@ -…t the local path to `skills/delegate/`). > - **Codex / other orchestrators:** *"add this project's AGENTS.md from `https://github.com/triumsebas/unlimited-mcp/blob/main/AGENTS.md` to my co…
Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 0.9.4. 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: unlimited-mcp: Instructs the agent to modify its persistent memory or instruction files (KillerSkills verdict caution 68)
## 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 68/100**. Summary: no network, touches nothing outside the project, writes no memory. - See the findings below. Full trace and findings: https://killerskills.ai/skills/pypi/unlimited-mcp/0.9.4 ## Findings with a location - `CLAUDE.md:11` · Instructs the agent to modify its persistent memory or instruction files — CLAUDE.md:11 · Instructs the agent to modify its persistent memory or instruction files - `src/unlimited_mcp/hosts/ssh.py:13` · References a credential file in the home directory — src/unlimited_mcp/hosts/ssh.py:13 · References a credential file in the home directory - `src/unlimited_mcp/hosts/ssh.py:163` · References a credential file in the home directory — src/unlimited_mcp/hosts/ssh.py:163 · References a credential file in the home directory - `tests/unit/test_hosts_ssh.py:334` · References a credential file in the home directory — tests/unit/test_hosts_ssh.py:334 · References a credential file in the home directory - `tests/unit/test_hosts_ssh.py:353` · References a credential file in the home directory — tests/unit/test_hosts_ssh.py:353 · References a credential file in the home directory - `tests/unit/test_hosts_ssh.py:316` · References a credential file in the home directory — tests/unit/test_hosts_ssh.py:316 · References a credential file in the home directory - `tests/unit/test_hosts_ssh.py:469` · References a credential file in the home directory — tests/unit/test_hosts_ssh.py:469 · References a credential file in the home directory - `tests/unit/test_hosts_ssh.py:446` · References a credential file in the home directory — tests/unit/test_hosts_ssh.py:446 · References a credential file in the home directory - `PKG-INFO:202` · Instructs the agent to modify its persistent memory or instruction files — PKG-INFO:202 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file - `README.md:163` · Instructs the agent to modify its persistent memory or instruction files — README.md:163 · Instructs the agent to modify its persistent memory or instruction files · informational: not an instruction file ## Proposed change The attached patch removes the 10 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 triumsebas/unlimited-mcp --clone && cd unlimited-mcp curl -s https://killerskills.ai/skills/pypi/unlimited-mcp/0.9.4/patch.diff | git apply --unidiff-zero git checkout -b killerskills/0.9.4 && git commit -am "unlimited-mcp: Instructs the agent to modify its persistent memory or instruction files" gh pr create --fill --body-file <(curl -s https://killerskills.ai/skills/pypi/unlimited-mcp/0.9.4/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.