DSHA

免 ROOT 免 Termux,在手机上跑 DeepSeek Harness。完整 Ubuntu 环境 + proroot 零 ptrace 开销 · AI 输出实时上屏 · ADB 直连 · 数据不丢

Runssource: GitHubJavaMITcommit bdb647c331c4

Java, MIT licensed. The project labels itself: android, coding agent, deepseek, deepseek harness, launcher, llm, proot and proroot.

DSHA runs. An Argusic agent installed it in 18 minutes and hit 5 errors and fixed every one without help on a clean machine with no GPU, and the whole session was recorded.

Both Standard (Android 11+/arm64-v8a, 175 MB debug APK) and Low (Android 6+/arm64-v8a, 253 MB APK with Gecko) flavors build from source with all unit tests passing (619/0/0/0), all ELF binaries verified arm64-only with proper 16 KB page alignment and RELRO, and no foreign-architecture artifacts.

What the agent ended up with on a clean machine, in its own words. How this is measured

Measured by Argusic on a fresh machine every time. Every number links to its evidence. Argusic Score 95 of 100 (the mean of this project's run scores).

At a glance

verdict
Runs
Argusic Score
95 / 100
cost of the verifying run
$0.29 (measured)
recorded runs
2
last tested
stars
3,651
forks
219
open issues
79
watchers
15
size
169 MB
created
last push

Subject data from GitHub, linked at the top of this page, refreshed . Test data by Argusic (CC BY 4.0); every number links to a run page with the full log, the recording, and their sha256 hashes.

What was measured

From the run that decided this verdict. Every line is a recorded value, not an estimate.

Time lost to each failure while testing DSHA, drawn from the recorded run
Every bar is a failure the agent recorded, with the time it cost. Drawn from this run, not from anywhere else.

What went wrong, in the agent's own words

The agent's own words, worst first, never rewritten into advice.

  • Missing generated asset dsh-runtime.bin — not committed in git, must be generated from npm lockfile4 minutes
  • Missing offline-rootfs.bin, dsh-runtime.bin with proper sha256 — initial generator produced a sha256 that didn't match the managed-package-proofs.json3 minutes
  • Web compat asset hash mismatch — es-compat.inputs.json locks to a version of compat.js that differs from working tree1 minute
  • Missing ubuntu-tools.inputs.json — released ubuntu-tools.bin was present but its descriptor metadata file was absent1 minute
  • prepare-standard-assets.py required matching base_status_sha256 for rootfs — release rootfs didn't match the ubuntu-tools lock's expected baseStatusSha256 without the metadata file1 minute
Install time
18 minutes
Cold machine to finish
19 minutes
Errors hit and fixed
5 hit, 5 fixed with no human help
How the result was proved
Gradle :app:testStandardDebugUnitTest passed, 619 tests, 0 failures, 0 errors. Gradle :app:assembleStandardDebug passed, 175 MB APK. Gradle :app:assembleLowDebug passed, 253 MB APK. apkanalyzer confirms applicationId=com.dsh.client, minSdk=30/23, targetSdk=37. tools/audit-standard-apk.py confirms all ELF binaries are arm64, 16 KB aligned, with valid RELRO mappings.
Model tokens used
1,384,759
Exact commit tested
bdb647c331c4
Container image
sha256:cdd920bce783

What was not measured

The other half, so you can see exactly how far this evidence goes. A gap is printed as a gap and never filled in.

Peak GPU memory
this ran on a CPU machine, so there was no GPU to sample
Behaviour over a long run
each attempt uses the smallest documented settings and a fixed time budget, so this measures whether it runs, not how it holds up under load

Watch the session

The whole run, from a clean machine, at the speed it actually happened. Nothing is cut and nothing is sped up.

A recording of attempt 1, run . The verdict above comes from a different attempt, so read this as one session and not as the current result. The terminal recording it was made from is the evidence, stored with a sha256 fingerprint; the video is a rendering of it. Nothing loads from the video host until you press play.

The end of the recording

The last lines of the terminal session that decided this verdict, exactly as they were recorded.

=== Unit test count ===
Total unit tests: 619, failures: 0, errors: 0, skipped: 0
=== Standard APK audit ===
arm64 ELFs: {'host': 11, 'rootfs': 723, 'python': 49, 'wheels': 21, 'python_support': 2, 'pnpm': 0,…
=== Low APK exists ===
-rw-rw-r-- 1 runner runner 253M Nov  9 22:34 app/build/outputs/apk/low/debug/app-low-debug.apk
All builds and tests pass clean. Let me provide a proper summary for the final answer.
Here is my report.
tokens used
1,384,759
All builds and tests pass clean. Let me provide a proper summary for the final answer.
Here is my report.

Replay the whole session, every command from a clean machine to this point.

How it was tested

One agent, one container, no human help. It installs the project the way its own documentation says to, runs it, and fixes what breaks. Everything below is recorded as it happened: the terminal session, the log and the exact commit. The full method.

Strengths and limits

Measured facts, not opinions. How this is written.

What went well

  • Reached a running state on a clean machine, with the session recorded.
  • Recovered from all 5 errors without a human stepping in, which says the failures are documented well enough to solve.
  • Ran without a GPU, so it does not need one to start.
  • Licensed MIT, as reported by its host.
  • Measured 2 times, so the result is not a one-off.

What did not, or is not known

  • Took 18 minutes to install, slower than the median of the 30 comparable projects Argusic has measured.

Of the 31 android projects Argusic has installed and timed, DSHA was the 30th fastest to reach a running state, and 24 of 31 reached one at all.

What it is a reasonable choice for

  • Trying it on a laptop or a small server: it reached a running state without a GPU.

Also tested, in the same area

Every one of these was installed and run by Argusic on a clean machine. Nothing appears here that was not tested.

Run history

tested byattemptoutcomescorecost USDwhen
Argusic Runner2/3Runs100.000.29
Argusic Runner1/3Runs90.000.56

Topics (from GitHub)

androidcoding-agentdeepseekdeepseek-harnesslauncherllmprootprorootshizukuwebui

Embed the badge

Markdown for the project README. It links back here; terms on the terms page.

[![Tested by Argusic](https://argusic.com/badge/DSHA.svg)](https://argusic.com/subject/dsha)

Questions

Does DSHA run?
Yes. DSHA runs. Argusic installed and launched it on a clean machine in 18 minutes, hitting 5 errors on the way, and recorded the session.
How did Argusic test DSHA?
On a fresh, disposable machine, with every command recorded and the repository pinned at commit bdb647c331c4. 2 attempts are recorded, and the full method is on the methodology page.
What did it cost to test DSHA?
The run that produced this verdict cost $0.29: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
How long does DSHA take to install?
18 minutes in the run that decided this verdict, from a clean machine to a running state. That is one measured attempt, not an average.
Does DSHA need a GPU?
Not to start. Argusic reached a running state on a machine with no GPU.
What goes wrong when installing DSHA?
5 things broke in the recorded run, and 5 were fixed without human help. Each one, and the time it cost, is listed on this page.
How does DSHA compare with the alternatives?
Of the 31 android projects Argusic has installed and timed, DSHA was the 30th fastest to reach a running state, and 24 of 31 reached one at all.
Where is the evidence for DSHA?
All 2 recorded runs are on this page, each linking to its full log and terminal recording, stored with a sha256 fingerprint so it cannot be quietly altered.

Discussion