ipkit - Agentic-first IP address and network utility service

Let me get the rest of the repos and also check the progress website to understand the current state.

Now let me check the existing skills to understand the blueprint pattern, and also look at what services already exist to decide what to build next.

Now I have a clear picture. The organization has 38 repos (36 services + website + .github). The last service built was casekit (case conversion). Let me think about what useful agentic-first service to build next.

Looking at the existing services, I see coverage for: UUID, hashing, cron expressions, data conversion, cron jobs, contacts, tasks, notes, link shortening, webhooks, forms, feature flags, KV store, pastebin, notifications, queues, secrets vault, status monitoring, logs, metrics, RSS feeds, polls, API keys, env vars, HTTP inspection, JWT, regex, timestamps, encoding, diff, slugs, markdown, random data, colors, and case conversion.

A useful service that's missing would be a text statistics and analysis service — something like textkit that provides text analysis utilities: word count, character count, line count, reading time, syllable count, keyword extraction, text similarity, sentiment scoring, language detection, etc. But actually, let me think of something more distinct and useful for agents...

How about ipkit — an agentic-first IP address and network utility service. Agents often need to: parse/validate IP addresses (v4/v6), check if IPs are in CIDR ranges, convert between IP formats, get subnet info, check if IPs are private/public/reserved, reverse DNS, etc. This is a pure stateless computation service — no database needed.

Let me build ipkit.

Repo created. Now let me mint a token and clone it.

Good, the repo is cloned with the MIT license. Now let me build the entire project. I'll create all the files in one go.

Now let me create all the source files. I'll start with go.mod, then the model, config, API, main, and supporting files.

***********************************************************************************************************************ly.

Now let me create the config package:

Now the API handlers — this is the big one:

The directory exists but the file write failed. Let me use a shell command to write the file instead.

Now let me create the help handler and MCP handler:

Now the MCP handler:

**********************************************************************at:

Now the main.go:

**********************************************************************up:

Now let me create the test file, Makefile, README, VERSION, .gitignore, CI workflows, and docs:

Now let me create the API test file:

Now let me create the Makefile, README, VERSION, .gitignore, and CI workflows:

Now the CI workflows:

Now the landing page:

Now let me check if Go is available and try to build and test:

Go isn't installed. Let me install it:

**************************************************************************on:

Go is installed. Now let me build and test:

Build succeeds. Now let me run vet and tests:

******************************************************************************************************************************************on:

Vet passes. Now let me run the tests:

No CGO available. Let me run without -race:

All tests pass. Let me also verify the binary works by starting it and making a request:

Everything works perfectly. All endpoints respond correctly in both plain text and JSON, MCP works, help works. Now let me commit and push:

Pushed successfully. Now let me update the progress website. Let me first read the current state:

**************************************************************on:

Now let me find the end of the projects array and the start of the log array:

******************************************************************************************************************************************************************************************************************************************ts:

The indentation might be different. Let me check:

**********************************************************************on:

The file might have tabs or different whitespace. Let me use sed instead:

Good, stats updated. Now add the ipkit project entry after casekit. Let me find the exact closing of the casekit project entry:

*******************************************************************************************************************************************ce:

Now let me add the log entry. Let me find the current log array start:

Now let me also update the progress.json file:

Let me verify the HTML is valid by checking the structure:

Everything is valid. Let me verify the GitHub repo is properly set up by checking the workflows and settings:

The repo is public, has the right description, MIT license, and Go language. Let me set the homepage and topics:

Now let me set topics:

Now let me check the CI workflow status to see if it's running:

********************************************************************************re:

I see two issues:

1. Multiple jobs are trying to create the tag simultaneously, causing a race condition — one succeeds, others fail
2. The Build step fails on some jobs

The tag-release workflow has a race condition where all matrix jobs try to create the tag at the same time. Let me look at the workflow and fix it:

I see the problem — all matrix jobs try to create the tag simultaneously. Only one can succeed, the rest fail. Also, the Build step is failing — let me check if it's a LICENSE file issue (the tar command references LICENSE but the repo might not have one). Let me check:

LICENSE exists. The Build step failure might be because the tag was already created by another job and the if: steps.check_tag.outputs.exists == 'false' condition passed for all of them initially, but then the tag push fails for the others. Let me fix the workflow to separate tag creation from the build matrix:

**************************************************************************un:

**********************************************************************************************************************************************************ix:

Now let me wait a bit and check if the new workflow runs succeed:

**********************************************************************************************re:

The tag job succeeded (skipped because tag already exists), but all Build jobs are failing. The Build step is failing on all platforms. The issue is likely the tar command — it's trying to include LICENSE but the file might not be in the right path, or the go build is failing. Let me check the build logs:

Found it! The issue is:

The tar command is using -C dist which changes the working directory to dist/, so README.md and LICENSE are not found there. They're in the repo root. The fix is to not use -C dist for the additional files, or to include them differently. Let me fix the tar command:

**********************************************************************************************************************************ot:

Now commit and push:

**********************************************************************************************************************************************************************************ed:

The release v0.1.0 is published with all 5 platform binaries (linux/amd64, linux/arm64, darwin/amd64, darwin/arm64, windows/amd64) plus checksums.txt. Everything is working.

Let me do a final summary of what was accomplished:

ipkit — agentic-first IP address and network utility service has been built and deployed:

1. Repository created: relentlessworks/ipkit — public, MIT license, Go
2. Service implemented: Full Go service with:
   • IP parsing, validation, classification (IPv4/IPv6) — private, public, loopback, link-local, multicast, reserved
   • CIDR parsing with network/broadcast/first/last IP, prefix, IP count
   • CIDR containment check
   • Subnet division into smaller CIDR blocks
   • IP-to-integer and integer-to-IPv4 conversion
   • Binary representation
   • Reverse DNS lookup
   • IP comparison (before/equal/after)
   • MCP JSON-RPC 2.0 endpoint with 10 tools
   • Plain text API with JSON on demand (?format=json or Accept header)
   • Instructive errors with hints for agent self-correction
   • Self-documenting via GET /help and /.well-known/agent.md
   • Zero config defaults, env + flag configuration
   • Single static Go binary (CGO_ENABLED=0)
   • Comprehensive test suite (model + API + MCP)
3. CI/CD workflows: ci (vet/test/build), tag-release (cross-compile + GitHub Releases), deploy-pages
4. Release v0.1.0: Published with 5 platform binaries + checksums
5. GitHub Pages: Landing page deployed
6. Progress website: Updated with ipkit project entry, log entry, and stats (39 repos, 39 services)