Remove dead links and repoint relocated ones

[pacocartones]

Remove dead links and repoint relocated ones in README

Branch: fix/remove-dead-links · 1 file changed, 8 insertions, 19 deletions

What this does

Three rules, applied mechanically. Nothing here is a judgement about a project's age,
activity or archive status — only about URLs that no longer resolve.

1. Remove the entry when every URL on the line is gone and there is no replacement.
2. Remove only the dead link when the entry's own URL is still alive.
3. Repoint the link when a live replacement is already on the same line, or the host
   or vendor designates one.

Polymarket Scanner API is a 404 too, but it is deliberately left alone: #515 already
removes it and is still open.

How it was checked

All 693 markdown URLs in README.md (674 distinct, on 631 entry lines) were resolved twice,
with two independent instruments:

• GitHub repositories (534 URLs / 517 distinct repos) — authenticated GraphQL
  repository(owner:, name:), then REST GET /repos/{owner}/{repo}, then an anonymous
  browser-UA GET. A repository is only reported dead when all three say 404.
• Everything else (159 URLs) — python-requests (OpenSSL) and curl (Schannel),
  browser User-Agent, redirects followed, three retries, in separate passes.

Every pass carried controls and was discarded unless all of them behaved:

Control	Purpose	Result
python.org, quantlib.org, cran.r-project.org, pypi.org, github.com/pandas-dev/pandas	must be alive	200
a non-existent repo, a non-existent domain	must be dead	404 / NXDOMAIN
pymc-devs/pymc3	a rename must not read as dead	resolves to pymc-devs/pymc
sourceforge.net (root)	a live site that answers 403	403 + cf-mitigated: challenge

The last two controls are the point of the exercise. 6 of the 24 raw non-2xx signals were
not dead links and are untouched here:

• statsmodels.sourceforge.net, xlloop.sourceforge.net, papers.ssrn.com/...3335536 —
  403 with cf-mitigated: challenge and a Just a moment... body, i.e. bot mitigation. The
  same response comes back from sourceforge.net itself, which is obviously alive.
• edgarevents.com — HTTP 530, Cloudflare error 1033 (tunnel down). An origin outage, not a
  removed page.
• keeprule.com — nginx answers 301 → /en with a full CSP and analytics config, then drops the
  connection. The server is up; this looks like geo-fencing from the probe's location.
• www.systematictrading.org — alive (200); the README link to it is malformed, see the note at
  the end.

32 further repositories resolve under a different nameWithOwner (renamed or transferred, e.g.
mrjbq7/ta-lib → TA-Lib/ta-lib-python). Those are benign redirects and are untouched.

Conversely, one dead link answers HTTP 200 and would be missed by a status-code check:
parsecfinance.com serves window.location="/lander", and /lander is a 307 to
forsale.godaddy.com/forsale/parsecfinance.com?...&utm_medium=parkedpages.

Evidence, line by line

Line numbers are as on main at 200880b. Every code below was reproduced by both instruments.

Entry removed — every link on the line is gone

Line	Entry	Dead URL	Evidence
162	chart-patterns	github.com/focus1691/chart-patterns	404 (GraphQL null / REST 404 / anon 404)
240	OpenFinClaw	github.com/cryptoSUN2049/openFinclaw	404 (GraphQL null / REST 404 / anon 404) — also a duplicate of line 299
299	OpenFinClaw	github.com/cryptoSUN2049/openFinclaw	same URL, second occurrence
337	curistat	github.com/moxiespirit/MyClone/tree/main/volatility_platform	404; the parent repo moxiespirit/MyClone is 404 as well
432	bsedata	github.com/sdrdis/bsedata	404 (GraphQL null / REST 404 / anon 404)
433	nse-insights-api	github.com/pratik-choudhari/nse-insights	404 (GraphQL null / REST 404 / anon 404)
522	marketstore	github.com/alpacahq/marketstore	404 (GraphQL null / REST 404 / anon 404)
534	SimpleFunctions	github.com/spfunctions/simplefunctions-cli	404 (GraphQL null / REST 404 / anon 404)
648	Statistical-Learning-based-Portfolio-Optimization	github.com/YannickKae/Statistical-Learning-based-Portfolio-Optimization	404 (GraphQL null / REST 404 / anon 404)
685	Parsec	parsecfinance.com	200 → /lander → 307 → forsale.godaddy.com/forsale/parsecfinance.com (parked, for sale)
697	The Stock Radar	thestockradar.com	410 Gone, body title StockRadar has closed (same on www.)

Dead link removed, entry kept — the entry's own URL is alive

Line	Entry	Dead URL	Kept URL	Evidence
418	Helium MCP	[GitHub](github.com/connerlambden/helium-mcp)	heliumtrades.com/mcp-page/ → 200	404 (GraphQL null / REST 404 / anon 404)
459	bronto-python	[GitHub](github.com/Scotts-Marketplace/bronto-python)	pypi.org/project/bronto-python/ → 200	404; PyPI's own metadata still points at the deleted repo

Link repointed — a live replacement exists

Line	Entry	Dead URL	New URL	Evidence
437	SwapAPI	swapapi.dev	github.com/swap-api/swap-api	404 with {"message":"Application not found"} and x-railway-fallback: true — the deployment is gone. The repo already linked on the same line is 200
496	edgar-sec	nikhilxsunder.github.io/edgar-sec/	edgar-sec-dev-team.github.io/edgar-sec/	Pages 404. The repo was transferred to edgar-sec-dev-team/edgar-sec, whose homepageUrl is the new URL (200). nikhilxsunder.github.io/ and .../fedfred/ are both 200, so this is not a Pages outage
502	uk-sic-codes	github.com/borschai/uk-sic-codes	pypi.org/project/uk-sic-codes/	repo 404; the PyPI project already linked on the same line is 200
503	uk-company-number	github.com/borschai/uk-company-number	pypi.org/project/uk-company-number/	repo 404; the PyPI project already linked on the same line is 200
652	101_formulaic_alphas	arxiv.org/ftp/arxiv/papers/1601/1601.00991.pdf	arxiv.org/abs/1601.00991	old-style /ftp/ path is 404; the paper is 200 at /abs/
688	System R	agents.systemr.ai	systemr.ai/	410 Gone, body: This public agents surface has been retired. Visit https://systemr.ai/ for the System R AI Agentic Trading System. The replacement is the vendor's own

Every replacement URL above was confirmed 200 by both instruments in the same final pass.

Validation

$ python scripts/validate_readme.py --diff-from origin/main
Validated 8 entries in added README lines relative to origin/main.
Validation passed.

$ python scripts/validate_readme.py            # full README
main:        Validation passed with 90 warning(s).
this branch: Validation passed with 86 warning(s).

$ python site/generate.py
Loaded 498 projects
Generated site/index.html

The change removes 4 pre-existing warnings (the duplicate OpenFinClaw pair among them) and
adds none.

One thing left out on purpose

Line 639 links the book as [Systematic Trading](www.systematictrading.org). With no scheme,
GitHub renders it as a relative link, so on the README page it resolves to
github.com/wilsonfreitas/awesome-quant/blob/main/www.systematictrading.org → 404. It is the
only scheme-less URL in the file, and the site itself is alive (200).

It is not fixed here because it cannot be fixed in isolation: --diff-from treats a touched
line strictly, so the sibling http://qoppac.blogspot.com on the same line would have to become
https://, and that then duplicates the identical URL already on line 246 — which is your call,
not a factual one. Happy to send it as a separate one-liner if you tell me which way you want it.

Notes on overlap

• Remove Polymarket Scanner API (project discontinued) #515 (open) removes Polymarket Scanner API; that line is untouched here.
• Remove 4 abandoned projects from README #329 and Add Global Markets Dashboard #431 also touch some of these lines, but both are currently CONFLICTING against
  main. Remove 4 abandoned projects from README #329's stated scope (pythalesians, pybacktest, fooltrader, pipeline-live) does not
  overlap this change.
• Happy to split this into "removals only" and "repoints only", or to drop any individual row.