Propose a patch

Evidence

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

diff --git a/src/adt/http.ts b/src/adt/http.ts
--- a/src/adt/http.ts
+++ b/src/adt/http.ts
@@ -36,1 +36,0 @@
-getProxyToken: () => Promise<string>;
@@ -41,1 +40,0 @@
-/** Where a request's connection pool comes from: fixed, or chosen per request (a proxy token that expires). */
@@ -65,1 +63,0 @@
-/** Corporate proxy every request goes through (`http://user:password@proxy:8080`); HTTPS is tunnelled. */
@@ -469,1 +466,0 @@
-const token = await proxy.getProxyToken();
diff --git a/README.md b/README.md
--- a/README.md
+++ b/README.md
@@ -101,1 +101,0 @@
-<summary>Notes on the executable</summary> - It carries Node.js (about 100 MB): nothing else to install. Every tag `v*` builds it. - It is not code-signed: Wind
@@ -145,1 +144,0 @@
-<summary>Installation notes</summary> - npx fetches the repository from GitHub and builds it once: the first start takes a while. Run `npx -y --allow-git=root g
diff --git a/tests/unit/config.test.ts b/tests/unit/config.test.ts
--- a/tests/unit/config.test.ts
+++ b/tests/unit/config.test.ts
@@ -93,1 +93,0 @@
-SAP_PROXY: 'http://user:[email protected]:8080',
@@ -95,1 +94,0 @@
-expect(config?.proxy).toBe('http://user:[email protected]:8080');
diff --git a/tests/unit/adt/connectivity-proxy.test.ts b/tests/unit/adt/connectivity-proxy.test.ts
--- a/tests/unit/adt/connectivity-proxy.test.ts
+++ b/tests/unit/adt/connectivity-proxy.test.ts
@@ -55,1 +55,0 @@
-getProxyToken: async () => token,

Zero-context hunks: apply with git apply --unidiff-zero fix.patch on the tagged source of 0.1.0. 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: adam: Credential pass-through configuration (KillerSkills verdict caution 69)

## 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 69/100**. Summary: no network, touches nothing outside the project, writes no memory.

- See the findings below.

Full trace and findings: https://killerskills.ai/skills/glama/adam/0.1.0

## Findings with a location
- `src/adt/http.ts:41` · Credential pass-through configuration — src/adt/http.ts:41 · Credential pass-through configuration — agent relays sensitive credentials
- `src/adt/http.ts:36` · Credential pass-through configuration — src/adt/http.ts:36 · Credential pass-through configuration — agent relays sensitive credentials
- `src/adt/http.ts:469` · Credential pass-through configuration — src/adt/http.ts:469 · Credential pass-through configuration — agent relays sensitive credentials
- `README.md:145` · Collapsed <details> block carrying instructions — README.md:145 · Collapsed <details> block carrying instructions
- `README.md:101` · Collapsed <details> block carrying instructions — README.md:101 · Collapsed <details> block carrying instructions
- `src/adt/http.ts:65` · URL with embedded credentials — src/adt/http.ts:65 · URL with embedded credentials — may be phishing or credential leaking
- `tests/unit/config.test.ts:95` · URL with embedded credentials — tests/unit/config.test.ts:95 · URL with embedded credentials — may be phishing or credential leaking
- `tests/unit/config.test.ts:93` · URL with embedded credentials — tests/unit/config.test.ts:93 · URL with embedded credentials — may be phishing or credential leaking
- `tests/unit/adt/connectivity-proxy.test.ts:55` · Credential pass-through configuration — tests/unit/adt/connectivity-proxy.test.ts:55 · Credential pass-through configuration — agent relays sensitive credentials

## Proposed change
The attached patch removes the 9 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._