Install verification: runtime verified (L5) · tested on dsh 0.2.1-rc.2 · last checked Install verification details →
What it doesAI
Adds a mood/persona layer to DeepSeek Harness with 33 evolving moods, a daily mood diary, cross-session memory, and news feeds.
- 33 moods with numeric values that drift every 3-6 hours
- Daily 23:59 mood diary, cross-session memory and taste library
- RSS/web/video news ingest with optional proactive speaking
Capability tags are AI-inferred or rule-extracted from the README (see badge) — vocabulary-limited, not install-verified. AI = inferred; README = literal keyword from the readme.
Installation
dsh plugin --profile web add github:JMX-4869/dsh-mood#v0.14.9Install method: GitHub · runtime verified (L5)
Verification & Compatibility
Grouped by plugin version: each card lists the dsh versions actually tested against that artifact version and their install verdicts (L5 wins). Untested combinations are omitted — absence means untested, never assumed compatible. L1–L3 are a plugin-level static archive (independent of the dsh version — the same verdict on every dsh row): they sit in each row's level grid next to L4/L5, marked "static", while their evidence is stored once per plugin inside that row's "checks & evidence" instead of being duplicated per version.
External credentials (plugin-level): none detected
○ The author declares no dsh version requirement — install is not blocked by version.
| plugin ＼ dsh | dsh 0.2.1-rc.2 current | dsh 0.2.1-alpha.2 |
|---|---|---|
| v0.14.9 | ✓ | · |
| v0.14.8 | · | · |
| v0.7.0 | · | ✓ |
v0.14.9github✓ 1 runtime-verified2026-10-21
https://github.com/JMX-4869/dsh-mood
✓dsh 0.2.1-rc.2current2026-10-21L5 · runtime verified
- • Node.js: v24.14.0
- • DSH: 0.2.1-rc.2
- • Plugin artifact: github · https://github.com/JMX-4869/dsh-mood#v0.14.9
ID: dsh-mood-2@v0.14.9@dsh0.2.1-rc.2 · profile: l3-fullscan
Issued: 2026-10-21 · expires: 2026-10-28
View checks & evidence
active in loader tree (150 entries)
L1–L3 are plugin-level static conclusions (baseline, identical across dsh versions): evidence is stored once per plugin.
v0.14.8github2026-10-18
https://github.com/JMX-4869/dsh-mood
No runtime install test for this version (static scan only) — compatibility unknown, never assumed.
v0.7.0github✓ 1 runtime-verified2026-10-15
https://github.com/JMX-4869/dsh-mood
✓dsh 0.2.1-alpha.22026-10-15L5 · runtime verified
- • Node.js: v24.14.0
- • DSH: 0.2.1-alpha.2
- • Plugin artifact: github · https://github.com/JMX-4869/dsh-mood#v0.7.0
ID: dsh-mood-2@v0.7.0@dsh0.2.1-alpha.2 · profile: l3-fullscan
Issued: 2026-10-15 · expires: 2026-10-22
View checks & evidence
active in loader tree (150 entries)
L1–L3 are plugin-level static conclusions (baseline, identical across dsh versions): evidence is stored once per plugin.
Only the latest 3 published versions are shown; source-snapshot (commit hash) groups, unstated groups and older versions are omitted — the full archive stays in the data layer (snapshot/unstated groups appear only when a plugin has no published version).
Author badges
Embed the official badges in your README to showcase security & install-test status:
`[](https://www.dsh.so/artifact/dsh-mood-2/)``[](https://www.dsh.so/artifact/dsh-mood-2/)`Security Report
Two automated layers, neither a manual review: a static scan (per-plugin rule checks) and a model-driven dynamic security audit that covers only the subset static-flagged at warn or above and never executes plugin code.
Static low, dynamic medium — stricter wins → overall medium.
v0.14.9dsh manifestv0.14.9git2026-10-21v0.14.12version moved onclean · gitVet is an AST static scan (npm artifact or git source).
show static findings (10)
- warningchildProcess · code-exec
asr.js:39const r = spawnSync(p, ['-c', 'print(1)'], { timeout: 8000, windowsHide: true, encoding: 'utf8' }) - warningchildProcess · code-exec
asr.js:50const r = spawnSync(py, ['-c', 'import faster_whisper, ctranslate2; print("ok")'], - warningchildProcess · code-exec
asr.js:332const r = spawnSync(ffmpegPath, ['-y', '-hide_banner', '-loglevel', 'error', - warningchildProcess · code-exec
measure.js:45const r = spawnSync(exe, [ - warningchildProcess · code-exec
measure.js:86const r = spawnSync(exe, ['-hide_banner', '-nostats', '-i', videoPath, '-vf', vf, - warningchildProcess · code-exec
measure.js:123const r2 = spawnSync(exe, ['-hide_banner', '-nostats', '-i', videoPath, '-af', 'silencedetect=n=-35dB:d=0.6', - warningfsWrite · files
memory.js:221if (lines.length) fs.appendFileSync(p, lines.join('\n') + '\n', 'utf8') - warningfsWrite · files
memory.js:260fs.writeFileSync(tmp, keep.map((o) => JSON.stringify(o)).join('\n') + (keep.length ? '\n' : ''), 'utf8') - warningfsWrite · files
memory.js:373fs.appendFileSync(to, a.endsWith('\n') ? a : a + '\n', 'utf8') - warningchildProcessImport · code-exec
scan-private.mjs:41import { execFileSync } from 'node:child_process'
show 3 audit findings
- mediumCross-session memory archive stores full chat text in plaintext on disk
memory.js:221Every message of 8+ characters (clipped at 8000) is appended verbatim as JSONL to a fixed, predictable path (~/.dsh/dsh-mood-memory.jsonl or a configured directory), shared across all sessions and kept until the 50 MB cap drops the oldest rows. Secrets typed into any chat (API keys, passwords, personal data) end up unencrypted in a home-directory file that the agent can search from any later session and that backups/cloud sync will copy. The sibling diary is encrypted with a key file, so the protection level here is inconsistent.
- lowGenerated Python script is written to a fixed shared temp path and then executed
asr.jsThe interpreter script always goes to os.tmpdir()/dsh-mood-asr/run.py (mkdirSync recursive, writeFileSync follows symlinks) and is then executed by python. On multi-user POSIX hosts a local attacker who pre-creates that directory or a run.py symlink can redirect the write into a file the victim may write, and the executed payload path is fully predictable (TOCTOU). Not reachable from remote/model input, so impact is limited to local attackers.
- lowWhisper model downloads silently defaulted to a third-party mirror (hf-mirror.com)
asr.jsWhenever transcription is enabled, faster-whisper resolves and downloads model weights through hf-mirror.com rather than huggingface.co, adding an unrequested third-party distribution channel for artifacts that are then loaded by ctranslate2. The user is not given an explicit choice, only a hint in an install string.
No RCE or backdoor; main issue is a plaintext cross-session chat archive on disk, plus low-risk fixed temp-script path and a default third-party model mirror.
Rule: the dynamic audit can only raise risk, never lower it (raise-only, never an exemption); a wrong raise is withdrawn by human review and the tier falls back to the static verdict. Overall verdict = the stricter of the two layers.
Disclaimer: the static scan and the dynamic security audit are both automated analysis, not a security guarantee. Always review what you install.
Activity
Last commit 2026-10-21 · activity: Active
• Repo created: 2026-10-13
• Stars: ★ 2 · Forks: 0
• Health: Active — committed within last 30 days
Source
GitHub: github.com/JMX-4869/dsh-mood
Author README
View on GitHub ↗The author's original README from the repository — fetched live, unedited (the English page loads the default README).