Skip to content

Releases

Atalaia 1.3.0

Choose a tag to compare

@jacksonfdam jacksonfdam released this 19 Aug 13:16
8967307

A report for the question the weekly digest does not answer — what the fleet is made of and how far behind it is — and the printing it turned out nobody had ever tried.

The dependency and technology report

GET /reports/dependencies, on the Reports page for the fleet and on each repository's own Report tab. One endpoint, one shape, two places to read it, so the two views cannot drift apart.

Four version states, never three. A dependency nobody has compared with a registry is unchecked — unknown, not up to date:

State Means
current The manifest already allows the newest published release
behind The registry has a release the manifest does not admit
unknown Checked and not comparable — a digest, a commit, an interval, or a registry that refused
unchecked Never compared with a registry

Folding unchecked into "current" is what would turn every percentage on the page into a claim nobody verified. On a real fleet the distinction is the whole report: 5857 dependencies came back as 129 current, 61 behind, 2 unknown and 5665 unchecked. So the report says how many were never asked, names the repositories nobody has scanned rather than counting them as clean, dates the newest registry answer it is reasoning from, and explains the pins it cannot compare instead of dropping them.

Packages behind are grouped across repositories and ranked by reach, because one upgrade that clears nine repositories is different work from nine unrelated ones. Technologies stay two independent signals — languages from the hosting provider describing the code, ecosystems from the manifests describing what it depends on — since a repository can report TypeScript and carry its risk in a Dockerfile.

Printing

Ctrl+P cut the page, and two of the causes lost content rather than looking wrong. .table-scroll scrolls sideways and the dependency tables cap their height inline; a printer does neither. Measured at a 700px page, the dependency table wanted 954px of width inside a 630px box and 1511px of height inside 384px — five columns and three quarters of the rows never reached the paper.

Print now has its own layout: one column, no navigation or buttons, nothing clipped, a fixed table layout so no column can push past the sheet, headers repeated on every page a table spills onto, and rows that do not split down the middle. The window titlebar was absolutely positioned and landed on top of the rows when a window broke across pages; in print it sits in the flow. Titlebars and table headers are white text on a coloured background, which disappears entirely when Chrome drops background graphics — its default — so they print black on white, and colour is forced only where it is the information: severity badges and the bars.

Fixes

  • The bar charts never had a fill. .bar-fill is a span, and width and height do not apply to an inline box, so every bar — Overview, feed health, the new report — measured 0x0 and drew an empty track. The track showed because it is a grid item and grid items are blockified; its child is not.

Verification

1268 tests across 69 suites, green. The route was exercised live in the running container (fleet 200, scoped 200, unknown id 404, non-integer id 400), and the print layout was measured and rendered at page width. Pagination itself was not confirmed visually: this machine has no PDF rasterizer, so header repetition and row splitting rest on standard table-header-group and break-inside: avoid rules rather than on a picture.

Full changelog: https://github.com/jacksonfdam/atalaia/blob/main/CHANGELOG.md

Atalaia 1.2.0

Choose a tag to compare

@jacksonfdam jacksonfdam released this 19 Aug 12:20
9d5bc00

Dependency coverage went from 14 of the 61 manifest and lock files people actually commit to 58, and the test suite stopped failing at random.

Full notes: CHANGELOG.md.

Lockfiles

Nothing was reading lockfiles. Every parser read a manifest, which declares a constraint: ^4.17.0 does not say whether the installed copy is 4.17.11, which is vulnerable to CVE-2021-23337, or 4.17.21, which is not.

Sixteen are read now — the npm family (package-lock.json schemas 1 and 2/3, pnpm-lock.yaml v5–v9, yarn.lock classic and Berry, bun.lock), Python's four, and composer.lock, Gemfile.lock, Cargo.lock, go.sum, packages.lock.json, paket.lock, gradle.lockfile, mix.lock, rebar.lock, pubspec.lock, Chart.lock, conan.lock, cabal.project.freeze, stack.yaml.lock, opam.locked, Cartfile.resolved and conda-lock.yml.

A lockfile now supersedes the manifest beside it. A repository holding both package.json and package-lock.json was storing lodash twice — once as ^4.17.0 and once as 4.17.21 — so every count doubled and the Dependencies tab listed the package twice with two different verdicts. Scoped by directory, so a monorepo keeps its per-app rows.

When a lockfile holds the same package twice, the lower version wins. rust-lang/cargo's own lockfile has 23 crates at two or three versions each. If 1.3.2 is vulnerable and 2.13.0 is patched, keeping the newer one turns a real exposure into a row that reads as clean.

New ecosystems

Dart and Flutter, Helm, Elixir and Erlang, C and C++ (Conan and vcpkg), Haskell, OCaml, Zig, Lua, Conda, Carthage, sbt and Ivy — with version lookups against pub.dev, hex.pm and Hackage.

Four ecosystems had been declared and never produced. The Helm one was visible: the console's Containers & CI only filter and the weekly digest both counted HELM, so they offered a category no row could ever match.

pyproject.toml

It was matched but barely read — one regular expression, and then every entry stored with a null version. Poetry below 2.0 writes no [project] table at all, so a Poetry project scanned as empty. Now: PEP 621, extras, PEP 735 groups as uv writes them, and the Poetry, PDM and Hatch tables, with PEP 503 name normalisation.

Sources and notifications

Debian security advisories, off by default. Read from the DSA/DLA lists rather than tracker/data/json — that endpoint is a status database, not a feed: 85.9 MB, 15.2 seconds, and no publication date anywhere, so every item would have been discarded.

A source that was never called no longer reports as empty. OpenCVE and CVE Details showed EMPTY with "Source responded but returned no vulnerabilities" — neither had responded. The status is NOT_CONFIGURED now, naming the setting.

NVD's 503 is a rate limit, not an outage. NVD_API_KEY is sent when set.

Discord, under Settings → Discord: an incoming webhook and an embed, truncated to Discord's limits because it rejects an over-long payload rather than trimming it.

The test suite

It was failing about one run in eight, on a different test each time. supertest starts a server for every single request, which across 1190 tests in eleven workers is thousands of bind-and-close cycles per run — and a port that has just been released still has the previous connection in TIME_WAIT. Each suite now binds one port and keeps it: 0 failures in 26 runs, against 3 in 8 before.

The migration advisory lock was keyed per database rather than per schema, so eleven workers queued twenty-two acquisitions on one key. The full suite went from 44–47 seconds to 21. That was the obvious suspect and it was not the cause.

Alerting

Old CVEs stopped being announced as new. CISA KEV serves its whole list on every fetch — 1670 items back to 2021 — and the feeds were substituting today's date for a missing one. Advisories are filtered by publication date, an undated one is discarded rather than dated to now, and a cycle sends at most 20 alerts with a pause between them.

Batch actions on the vulnerabilities table, model answers stored as answers (no preamble, no fences, no restated titles), and Ollama asked for the answer rather than the reasoning — think: false took a 12B model from 57.7 seconds to 6.2 for the same paragraph.

Verification

1229 tests, 67 suites
26 consecutive clean full runs

Atalaia 1.1.0

Choose a tag to compare

@jacksonfdam jacksonfdam released this 16 Aug 16:39
940643a

First tagged release; main carried no tag until now.

MCP server

POST /mcp — twelve tools for agents: search CVEs, open one in full with the repositories and owners it touches, list a repository's dependencies and what is behind, read the weekly digest. All read-only except explain_vulnerability, which asks the configured model. Stateless transport, behind the same API key as the REST API.

Telegram

Cycle alerts with Acknowledge and Resolve buttons, a weekly digest of outdated dependencies, and direct messages to system owners. The bot answers with the chat id of whoever writes to it — the one piece of configuration nobody can look up — and ignores anyone who is not the configured destination.

Tunnels

A registry with one file per provider: ngrok and Cloudflare quick tunnels. PUBLIC_URL beats both. atalaia.sh up prints the public address it managed to get, and the console's General page shows it.

Documentation site

docs/ renders as a static site carrying the console's own chrome. Navigation comes from the index table in docs/README.md, so a page that is not listed does not exist.

Fixes

  • The production image was missing package.json; the API container crash-looped at boot while the worker stayed healthy, so the stack looked half-alive.
  • Wiki pointers removed from the documentation.
  • Ecosystem count corrected: twelve, not fourteen.