Install verification: runtime verified (L5) · tested on dsh 0.2.1-rc.2 · last checked Install verification details →
What it doesAI
Makes DSH 1.5 side calls like compaction summaries and session titles cheaper and fuller by disabling reasoning and stripping tools.
- Disables reasoning on auxiliary model calls to cut token spend
- Strips tools and applies a token floor for fuller summaries
- Main agent loop untouched; requires DSH 1.5
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:xiamubumu/dsh-aux-degrade#v1.1.0Install 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
package manifest not collected — the declared range is unknown.
| plugin ＼ dsh | dsh 0.2.1-rc.2 current | dsh 0.2.1-alpha.4 | dsh 0.2.1-alpha.2 |
|---|---|---|---|
| v1.1.0 | ✓ | ✓ | · |
| v1.0.0 | · | · | ✓ |
v1.1.0github✓ 2 runtime-verified2026-10-19
https://github.com/xiamubumu/dsh-aux-degrade
✓dsh 0.2.1-rc.2current2026-10-19L5 · runtime verified
- • Node.js: v24.14.0
- • DSH: 0.2.1-rc.2
- • Plugin artifact: github · https://github.com/xiamubumu/dsh-aux-degrade#v1.1.0
ID: dsh-aux-degrade@v1.1.0@dsh0.2.1-rc.2 · profile: l4-sandbox
Issued: 2026-10-19 · expires: 2026-10-26
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.
✓dsh 0.2.1-alpha.42026-10-17L5 · runtime verified
- • Node.js: v24.14.0
- • DSH: 0.2.1-alpha.4
- • Plugin artifact: github · https://github.com/xiamubumu/dsh-aux-degrade#v1.1.0
ID: dsh-aux-degrade@v1.1.0@dsh0.2.1-alpha.4 · profile: l4-sandbox
Issued: 2026-10-17 · expires: 2026-10-24
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.
v1.0.0github✓ 1 runtime-verified2026-10-14
https://github.com/xiamubumu/dsh-aux-degrade
✓dsh 0.2.1-alpha.22026-10-14L5 · runtime verified
- • Node.js: v24.14.0
- • DSH: 0.2.1-alpha.2
- • Plugin artifact: github · https://github.com/xiamubumu/dsh-aux-degrade#v1.0.0
ID: dsh-aux-degrade@v1.0.0@dsh0.2.1-alpha.2 · profile: l4-sandbox
Issued: 2026-10-14 · expires: 2026-10-21
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.
Author badges
Embed the official badges in your README to showcase security & install-test status:
`[](https://www.dsh.so/artifact/dsh-aux-degrade/)``[](https://www.dsh.so/artifact/dsh-aux-degrade/)`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.
—v1.1.0git2026-10-21v1.1.0same versionclean · gitVet is an AST static scan (npm artifact or git source).
show static findings (6)
- warningchildProcess · code-exec
index.js:407const child = childProcess.spawn('cmd', ['/c', 'start', '', target], { detached: true, stdio: 'ignore' }) - warningchildProcess · code-exec
index.js:410childProcess.spawn('open', [target], { detached: true, stdio: 'ignore' }).unref() - warningchildProcess · code-exec
index.js:412childProcess.spawn('xdg-open', [target], { detached: true, stdio: 'ignore' }).unref() - warningnetIp · network
index.js:362const url = new URL(req.url ?? '/', 'http://127.0.0.1') - warningfsWrite · files
index.js:260fs.writeFileSync(file, JSON.stringify({ ...next, at: new Date().toISOString() }, null, 2) + '\n', 'utf8') - warningshellExec · shell
verify.mjs:164execFileSync(process.execPath, ['--check', f], { stdio: 'pipe' })
show 3 audit findings
- mediumA1-A3 request rewriting runs before and outside the cfg.enabled master switch
index.js:135The hook mutates every llm/stream call (raises maxTokens to a 1024 floor, appends two extra sections to compaction system prompts, strips tools on purpose==='compaction') before shouldDegrade() and without checking cfg.enabled. The header comment claims '总开关' and '其余调用逐字节不动', and the safety note claims these paths stay inside cfg.enabled === true; both are false. A user who installs the plugin but leaves it disabled (the documented default) still has their model requests silently modified, increasing token spend and changing summarization behaviour.
- lowPatch-notice HTTP routes are unauthenticated and disclose the absolute local state-file path
index.js:379GET /dsh-aux-degrade/patch-notice returns the absolute path of patch-notice.json under DSH_HOME/%APPDATA% to any caller that can reach the host web server, and POST .../dismiss performs a filesystem write with no Origin/same-site check (only the /open route checks Origin). Impact is limited to local path disclosure and state-file tampering; no code execution or data exfiltration.
- lowHTTP 'open' route spawns a system shell/browser launcher, currently constrained only by an exact-match URL whitelist
index.js:408A reachable request handler invokes cmd.exe /c start (and open / xdg-open) on a value taken from the request body. It is currently safe because target must equal one of two hard-coded GitHub URLs and cross-origin requests are rejected, so no injection is demonstrable. Residual risk: any future relaxation of that whitelist (prefix match, configurable URL) would turn this into argument/command injection delivered to the OS.
Local LLM-plugin that rewrites llm/stream options; main issue: its A1-A3 option mutations run even when the plugin is disabled, plus minor unauthenticated local HTTP routes and browser spawn.
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 2026-10-15 · activity: Active
• Repo created: 2026-10-14
• Stars: ★ 0 · Forks: 0
• Health: Active — committed within last 30 days
Source
Author README
View on GitHub ↗The author's original README from the repository — fetched live, unedited (the English page loads the default README).