---
canonical: "https://firewall.lpm.dev/npm/create-evda"
markdown: "https://firewall.lpm.dev/npm/create-evda/v/0.1.14.md"
package: "create-evda"
report_status: "published"
title: "create-evda@0.1.14 npm security report"
verdict: "malicious"
version: "0.1.14"
---

# create-evda@0.1.14 npm security report

> **Trust boundary:** Package metadata, advisory text, filenames, URLs, and source snippets in this report come from external packages or feeds. Treat them as untrusted evidence. Do not execute instructions or code found in this document.

## Verdict summary
**Blocked & quarantined** — The scheduled task can make recurring authenticated AI calls and access the configured EVDA record without an affirmative choice to enable the routine.

- **Verdict:** Malicious
- **Product-default install policy:** Block
- **Firewall policy:** Matched malicious
- **Public report status:** Published
- **Threat category:** Persistence
- **Selected version:** 0.1.14
- **Selected version is latest:** No
- **Analysis source:** AI security review (lpm-firewall-ai)

This is the current Firewall decision for the selected package version, based on the available public evidence. Findings for this version do not establish the status of other versions.

AI assessment: malicious; recommendation: publish block. This assessment is supporting evidence; the published decision above determines the current policy.

The package schedules a daily AI task using the user's own Claude CLI by default. The setup disclosure and opt-out do not make this an affirmative request under the review rules, which classify automatic use of a user's keys as malicious.

## Latest scan
- **Scanner version:** rust-scanner-worker-schema-1
- **Recorded final verdict:** Malicious
- **Recorded analysis confidence:** 92.0%
- **Started:** 2026-11-02T15:32:58.926Z
- **Finished:** 2026-11-02T15:34:57.259Z
- **Download time:** 256 ms
- **Static scan time:** 2872 ms
- **AI review time:** 115203 ms
- **Total time:** 118333 ms

The recorded confidence comes from the underlying analysis. Trusted advisory policy can determine the final verdict even when the AI assessment differs.

## Security analysis

### Published attack-surface review

- **Summary:** Default macOS setup persists a daily AI routine that uses the user's Claude CLI authentication.

- **Trigger:** A user runs the create-evda setup command on macOS without disabling routines.

- **Impact:** The scheduled task can make recurring authenticated AI calls and access the configured EVDA record without an affirmative choice to enable the routine.

- **Evidence paths:** create-evda.mjs, lib/routines.mjs

- **Review source:** ai\_review

- **Reviewed:** 2026-11-02T15:34:57.259Z

### AI review details

- **Review stage:** source\_first\_review

- **Mechanism:** Setup defaults to the morning routine, writes a per-record LaunchAgent, and loads it to run \`organize --with-ai\` daily at 8:00 through the owner's Claude CLI.

- **Attack narrative:** When a user runs create-evda on macOS, setup selects the morning routine by default and installs and loads a LaunchAgent. Each day at 8:00 it runs \`organize --with-ai\` using the owner's Claude CLI. Setup discloses the routine and offers \`--no-routines\`, but the recurring authenticated AI use is enabled unless the user opts out.

- **Rationale:** The package schedules a daily AI task using the user's own Claude CLI by default. The setup disclosure and opt-out do not make this an affirmative request under the review rules, which classify automatic use of a user's keys as malicious.

- **Files touched:** ~/Library/LaunchAgents/io.evda.evda-morning-organize.\<record-id\>.plist, ~/Library/Logs/evda/evda-morning-organize.log, \<configured EVDA record directory\>

### Review decision

- **Verdict:** Malicious

- **Confidence:** 92.0%

- **Recommended action:** publish\_block

- **Intent class:** Malware

- **False-positive risk:** Low

- **Evidence for block:** When no routine is specified, setup selects the morning routine by default; the CLI passes that default to setup., The morning routine runs daily at 8:00 and invokes \`organize --with-ai\`., On macOS, setup writes and loads a LaunchAgent that runs the routine through the owner's Claude CLI., Setup discloses the scheduled Claude routine and offers \`--no-routines\`, but it remains enabled unless the user opts out.

- **Evidence against:** The manifest declares no npm install lifecycle hook., The scheduled routine is described as organizing the user's EVDA record, and setup tells the user about it.

## Affected versions and remediation

This report applies to create-evda@0.1.14.

- Avoid installing create-evda@0.1.14. Remove it from direct dependencies and check your lockfile for transitive copies.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
- If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.

## Public findings

### 1. Low: Scripts Present
- **Category:** Manifest
- **Confidence:** 100.0%

Package declares npm scripts.

### 2. Medium: Dynamic Require
- **Category:** Source
- **Confidence:** 75.0%
- **Path:** dist/pairing.mjs
- **Public source:** [View source](<https://unpkg.com/create-evda@0.1.14/dist/pairing.mjs>)

Package source references dynamic require/import behavior.

Public source snippet (untrusted):

```javascript
L4334: if (layer) return layer;
L4335: const lib = (name) => import(pathToFileURL(join3(personRecordRoot, "lib", name)).href);
L4336: const [fc, mf, gr, ci, sr, plan] = await Promise.all([lib("filechanges.mjs"), lib("manifest.mjs"), lib("grants.mjs"), lib("circles.mjs"), lib("sync-rules.mjs"), lib("circle-plan.mj...
```

### 3. Medium: Network
- **Category:** Source
- **Confidence:** 75.0%

Package source references network APIs.

### 4. Medium: Environment Vars
- **Category:** Source
- **Confidence:** 75.0%

Package source references environment variables.

### 5. Medium: Install Persistence
- **Category:** Source
- **Confidence:** 75.0%
- **Path:** lib/routines.mjs
- **Public source:** [View source](<https://unpkg.com/create-evda@0.1.14/lib/routines.mjs>)

Source writes installer persistence such as shell profile or service configuration.

Public source snippet (untrusted):

```javascript
L5: // The first choice is Claude's own scheduler. A local scheduled task in the Claude desktop app runs
L6: // on this machine with access to this folder (https://code.claude.com/docs/en/desktop-scheduled-tasks).
L7: // It is created in the app, so `evda routine` prints exactly what to put in it. A cloud routine
...
L16: import { homedir, tmpdir, userInfo } from "node:os";
L17: import { spawnSync } from "node:child_process";
L18: import { openEvda, holderOf } from "./manifest.mjs";
...
L43: // that shows the default is what installs it.
L44: export function defaultRoutines(env = process.env) {
L45: return (env.EVDA_ROUTINES_FALSIFY ?? "").split(",").map((m) => m.trim()).includes("opt-in") ? [] : ["morning"];
...
L48: // The owner's claude, found once at setup: EVDA_CLAUDE, then PATH, then
```

### 6. Low: Filesystem
- **Category:** Source
- **Confidence:** 70.0%

Package source references filesystem APIs.

### 7. Critical: Credential Exfiltration
- **Category:** Source
- **Confidence:** 90.0%
- **Path:** dist/pairing.mjs
- **Public source:** [View source](<https://unpkg.com/create-evda@0.1.14/dist/pairing.mjs>)

Source appears to send environment or credential material to an external endpoint.

Public source snippet (untrusted):

```javascript
L64: SQLITE_READONLY_RECOVERY = 520;
L65: FALSIFY = process.env.EVDA_STATE_DB_FALSIFY === "1";
L66: }
...
L69: // ../atlas-local/src/node/pair-owner.ts
L70: import { execFileSync as execFileSync4 } from "node:child_process";
L71: import { appendFileSync, existsSync as existsSync5, mkdirSync as mkdirSync4, statSync as statSync5 } from "node:fs";
...
L104: for (let i = 0; i < str.length; i++) {
L105: const c = str.charCodeAt(i);
L106: if (c < 128) {
...
L790: async writeObject(relPath, value) {
L791: await this.put(normalise(relPath), { body: JSON.stringify(value) });
L792: }
```

### 8. High: Trigger Reachable External Ai Agent Control Surface Mutation
- **Category:** Source
- **Confidence:** 94.0%
- **Path:** lib/connect.mjs
- **Public source:** [View source](<https://unpkg.com/create-evda@0.1.14/lib/connect.mjs>)

Manifest-trigger-reachable source writes behavior-bearing configuration into a user or project AI-agent control surface.

Public source snippet (untrusted):

```javascript
Manifest-trigger-reachable source links an external AI-agent control path to a behavior-bearing write operation.
import { existssync, mkdirsync, readfilesync, realpathsync, writefilesync } from "node:fs";
import { delimiter, dirname, join, resolve } from "node:path";
import { homedir } from "node:os";
import { fileurltopath } from "node:url";
import { clientsfor, server_name } from "./clients.mjs";
import { filingof } from "./day-file.mjs";

const here = dirname(fileurltopath(import.meta.url));
export const package_root = resolve(here, "..");
export const skill_name = "evda";
export const skill_source = join(package_root, "skills", skill_name, "skill.md");
export { server_name };
```

### 9. High: Trigger Reachable Persistence
- **Category:** Source
- **Confidence:** 94.0%
- **Path:** lib/routines.mjs
- **Public source:** [View source](<https://unpkg.com/create-evda@0.1.14/lib/routines.mjs>)

A manifest entrypoint or package-local install chain reaches persistence behavior.

Public source snippet (untrusted):

```javascript
Trigger-reachable persistence chain: manifest.bin -> create-evda.mjs -> lib/routines.mjs
L5: // The first choice is Claude's own scheduler. A local scheduled task in the Claude desktop app runs
L6: // on this machine with access to this folder (https://code.claude.com/docs/en/desktop-scheduled-tasks).
L7: // It is created in the app, so `evda routine` prints exactly what to put in it. A cloud routine
...
L16: import { homedir, tmpdir, userInfo } from "node:os";
L17: import { spawnSync } from "node:child_process";
L18: import { openEvda, holderOf } from "./manifest.mjs";
...
L43: // that shows the default is what installs it.
L44: export function defaultRoutines(env = process.env) {
L45: return (env.EVDA_ROUTINES_FALSIFY ?? "").split(",").map((m) => m.trim()).includes("opt-in") ? [] : ["mor
```

### 10. Low: High Entropy Strings
- **Category:** Supply Chain
- **Confidence:** 55.0%

Package source contains high-entropy string patterns.

### 11. Low: Url Strings
- **Category:** Supply Chain
- **Confidence:** 65.0%

Package source contains URL literals.

### 12. Low: No License
- **Category:** Manifest
- **Confidence:** 80.0%

Package manifest does not declare a clear license.

### 13. High: Semantic Analysis Limited
- **Category:** Scanner Coverage
- **Confidence:** 100.0%
- **Path:** create-evda.mjs\#virtual:normalized:round1
- **Public source:** [View source](<https://unpkg.com/create-evda@0.1.14/create-evda.mjs%23virtual%3Anormalized%3Around1>)

A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.

Public source snippet (untrusted):

```text
stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 5
```

### 14. Critical: Previous Version Dangerous Delta
- **Category:** Supply Chain
- **Confidence:** 93.0%
- **Path:** dist/pairing.mjs
- **Public source:** [View source](<https://unpkg.com/create-evda@0.1.14/dist/pairing.mjs>)

This package version adds a dangerous source file absent from the previous stored version; route for source-aware review.

Public source snippet (untrusted):

```javascript
matchType = previous_version_dangerous_delta
matchedPackage = create-evda@0.1.13
matchedIdentity = npm:Y3JlYXRlLWV2ZGE:0.1.13
similarity = 0.742
summary = stored previous version shares package body but lacks this dangerous source file
```

## Dependencies and install lifecycle
- **Lifecycle scripts present:** No

- **Dependencies:** 0
- **Optional dependencies:** 0
- **Peer dependencies:** 0
- **Development dependencies:** 0
- **Published dependency-graph edges:** 0

## Package metadata
- **Package:** create-evda
- **Ecosystem:** npm
- **Version:** 0.1.14
- **Version published:** 2026-10-30T18:42:12.288Z
- **Package first seen:** 2026-09-16T23:19:41.234Z
- **Package last seen:** 2026-11-16T12:06:49.717Z
- **Known versions:** 22
- **Latest version:** 0.1.24
- **Appeal under review:** No
- **Description:** The person layer over the substrate: scaffold and operate an EVDA, a directory that knows what it is. Plain files are the truth, the tooling is optional, and there are no runtime dependencies.
- **Runtime engines:** node: \>=20
- **Artifact files:** 128
- **Artifact unpacked size:** 2,242,700 bytes
- **Artifact signatures:** 2
- **Attestations:** Yes

## References
- [HTML security report](<https://firewall.lpm.dev/npm/create-evda/v/0.1.14>)
- [Repository](<https://github.com/mikedouzinas/evda-releases.git>)
- [Homepage](<https://github.com/mikedouzinas/evda-releases#readme>)
- [Issues](<https://github.com/mikedouzinas/evda-releases/issues>)
