agent-safe-pipeline

Open-source gateway that puts authorization in front of every action your AI agents, workflows and pipelines take, before it commits: payments, trades, price changes, deployments and API calls run once on a single-use grant, wait for a person, or are refused, as Decionis decides, with chained evidence. Homebrew, Linux, Docker, Helm.

Runssource: GitHubhomepageTypeScriptApache-2.0commit 2126d70fb4ce

TypeScript, Apache-2.0 licensed. The project labels itself: agent security, ai agents, ai governance, api gateway, audit trail, authorization, aws and azure.

agent-safe-pipeline runs. An Argusic agent installed it in 14 minutes and hit 3 errors and fixed every one without help on a clean machine with no GPU, and the whole session was recorded.

The AgentSafe repo installs and builds cleanly under user-land Node 22.14.0 + Go 1.26.4; the boundary test, smoke test, all 6 workspace test packages, the live enforcement gateway (ALLOW 201 forwarded, ESCALATE 202 held, BLOCK 403 refused), chain verification, and conformance verifier all pass, and the standalone executable answers version 0.2.5.

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 100 of 100 (the mean of this project's run scores).

At a glance

verdict
Runs
Argusic Score
100 / 100
cost of the verifying run
$0.10 (measured)
recorded runs
1
last tested
stars
661
forks
57
open issues
15
watchers
4
size
15 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 agent-safe-pipeline, 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.

  • System node is v18.19.1 but the repo requires Node >=22.14.0 (and pnpm/corepack were absent)4 minutes
  • The Go toolchain needed by the govern contract test (go build ./cmd/govern) was missing3 minutes
  • The recursive workspace test run initially failed at @decionis/govern-contract with ERR_TEST_FAILURE/ERR_ASSERTION because the Go binary could not be built
Install time
14 minutes
Cold machine to finish
16 minutes
Errors hit and fixed
3 hit, 3 fixed with no human help
How the result was proved
pnpm -r --workspace-concurrency=1 test exited 0 with 145+275+836+6+12+17 tests passing; agentsafe test printed Verdict BOUNDARY HOLDS (exit 0); curl -X POST to the live gateway returned 201/202/403 for ALLOW/ESCALATE/BLOCK; packaging/smoke/Smoke.sh printed SMOKE PASSED; verify chain returned ok:true
Model tokens used
491,031
Exact commit tested
2126d70fb4ce
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

The end of the recording

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

- Read `README.md` and `docs/quickstart/README.md`; this is the AgentSafe monorepo (Node ≥22.14, pn…
- Container had only Node 18 (too old) and no pnpm/go. Installed Node 22.14.0 and Go 1.26.4 as user…
- `pnpm install --frozen-lockfile` succeeded (exit 0); built both core packages per the quickstart;…
- Launched the real gateway (`agentsafe proxy --upstream http://localhost:3000 --port 8080`) agains…
- Ran `agentsafe test` → `Verdict BOUNDARY HOLDS`, exit 0, 26 chained lines verified. `smoke.sh` → …
- Full workspace suite `pnpm -r --workspace-concurrency=1 test` exits 0: commerce 10 files/145 test…
**Errors hit (all fixed)**
1. Node 18 too old → installed Node 22.14.0.
2. Corepack couldn't create `~/.local/bin` (missing dir) → `mkdir -p` then enable.
3. pnpm bin-links warned (dist not built) → benign; built packages.
4. `govern` contract test failed (`go: not found`, hook failed 0/6) → installed Go 1.26.4; then 6/6…
5. My own shell mistakes (background process died with shell; wrong env var for journal dir; mixed …

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.
  • Installed in 14 minutes.
  • Recovered from all 3 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 Apache-2.0, as reported by its host.

What did not, or is not known

  • Measured once so far. A single attempt shows it can run, not that it runs reliably.

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 Runner1/3Runs100.000.10

Topics (from GitHub)

agent-securityai-agentsai-governanceapi-gatewayaudit-trailauthorizationawsazuredockerexecution-authorityhelm-charthuman-in-the-loopistiokubernetesmcpon-premisepolicy-as-codereverse-proxyself-hostedservice-mesh

Embed the badge

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

[![Tested by Argusic](https://argusic.com/badge/agent-safe-pipeline.svg)](https://argusic.com/subject/agent-safe-pipeline)

Questions

Does agent-safe-pipeline run?
Yes. agent-safe-pipeline runs. Argusic installed and launched it on a clean machine in 14 minutes, hitting 3 errors on the way, and recorded the session.
How did Argusic test agent-safe-pipeline?
On a fresh, disposable machine, with every command recorded and the repository pinned at commit 2126d70fb4ce. One attempt is recorded, and the full method is on the methodology page.
What did it cost to test agent-safe-pipeline?
The run that produced this verdict cost $0.10: real compute and model cost, not a list price per million tokens. It is recorded on that run's page.
How long does agent-safe-pipeline take to install?
14 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 agent-safe-pipeline need a GPU?
Not to start. Argusic reached a running state on a machine with no GPU.
What goes wrong when installing agent-safe-pipeline?
3 things broke in the recorded run, and 3 were fixed without human help. Each one, and the time it cost, is listed on this page.
Where is the evidence for agent-safe-pipeline?
The recorded run is on this page, each linking to its full log and terminal recording, stored with a sha256 fingerprint so it cannot be quietly altered.

Discussion