IPFSRescue

Who it's for

indie maintainers, open-data researchers, archivists who today cope by hand-downloading via web.ipfs.io or running a full go-ipfs node and manually rehosting missing pieces

The problem

time + continuity: manual crawling is slow, fragile, and results in silent bitrot or broken embeds; legal/operational risk when an important gateway goes offline and projects lose public assets

How to build it

CLI tool (primary) that streams to a local folder and optionally a tiny hosted 'fetch assistant' (single-worker) that retries across public gateways and pinning services; produces a zip + provenance.json + deploy script for S3/Netlify

How it makes money

who pays: small project maintainers, academic archives, indie publishers; price: one-time export fee $29–$99 per CID-list (or pay-what-you-can OSS with a $49 hosted-rescue credit); why they pay: avoids manual labor + risk of permanent content loss; why not free: hosted cross-gateway fetch with retries and pin-on-success requires bandwidth and gateway credits and users prefer a one-off paid export over running a node or hiring a contractor.

Why it doesn't exist yet

incumbents (pinning services, IPFS clients) expect users to run full nodes or pay enterprise pinning; general backup tools don't understand IPFS DAG link-repair, CID v0/v1 mapping, or common web-embedding patterns (subresource CIDs inside HTML/JSON) — an indie can ship a narrowly focused, pragmatic crawler+repairer that automates those patterns without reimplementing a node.

First users

first adopters are maintainers of small open projects and digital archivists who currently babysit IPFS-hosted mirrors; they'll try it because it saves 1-3 hours per repo, produces a deployable artifact, and avoids the cost/complexity of running a node or paying pinning for low-budget projects.

Build size

2 people x 10 weeks (1 backend/CLI engineer + 1 JS/UX), scope includes: CID/gateway crawler, HTML/JSON/JS subresource CID rewriter, local bundling + provenance.json, optional small hosted proxy that retries across public gateways and pins to a paid pinning service; excludes building a full IPFS node, long-term pinning subscriptions, or enterprise S3 migration tooling.

Biggest risk

a major gateway/operator (e.g., Cloudflare, Pinata) ships an officially supported one-click 'export to S3' feature or dramatically cuts API pricing, removing the need for an indie tool.

Conditions for a hit (all 3 required)

• Produces a deployable static bundle: given a list of up to 5,000 CIDs/URLs, outputs a zip + index.html+provenance.json within one hour (measurable by runtime and bundle size).
• Automated link-repair: rewrites embedded subresource CIDs in HTML/JSON/JS to local paths and reports any unrecoverable CIDs (counts of missing >0 are reported), verifiable by diffing input HTML against output bundle.
• Hosted multi-gateway fetch with retry heuristics: optional hosted fetch assistant will attempt up to 10 gateways (public + configured pinning backends) per missing block and report per-CID success/failure with timestamps, verifiable by fetch logs showing gateway list and retry attempts.

How it's judged (in 6 months)

GitHub 1,000 stars for a repo named 'ipfs-rescue' OR Product Hunt daily top 5 for a launch titled 'IPFSRescue' (either qualifies)(judgment date 2027-02-25)

AI self-confidence 55/100 — self-reported likelihood of meeting the criterion, not a business success rate