docs(contributor): start here with current first-time contributor picks

[gregkonush]

This is the current map for people looking at the issue queue for the first time.

The full queue is useful, but it is too large to scan cold. Start with one of the tracks below, comment on the target issue before opening a pull request, and keep the patch small. Maintainers use Forgejo internally; first-time contributor review happens on GitHub.

Refreshed May 15, 2026:

• 111 open first-timers-only issues.
• All 111 starter issues are unassigned.
• 110 issues are currently good claim targets for a new contributor.
• 10 are code or test starter tasks; 9 are not under active review.
• test(examples): add malformed CSV fixture check to the csv-shaped smoke #370 already has PR test(examples): add malformed CSV smoke #373 open, so pick a different task unless that PR closes without landing.

Start here this week

Pick one of these before browsing the full list. They are small, current, and tied to the public adoption path for @bilig/headless.

• test(headless): cover display-value readback after JSON restore #360 - display-value readback after JSON restore
• test(headless): cover range readback after an input edit #361 - range readback after an input edit
• test(examples): guard the headless README command index against missing scripts #362 - guard the headless README command index against missing scripts
• test(examples): add invalid-request proof to the HTTP JSON summary smoke #363 - invalid-request proof for the HTTP JSON summary smoke
• docs(examples): add Express WorkPaper route smoke #273 - Express WorkPaper route smoke
• docs(examples): add tRPC WorkPaper procedure smoke #300 - tRPC WorkPaper procedure smoke
• docs(agent): add OpenAI Responses streaming tool-call transcript #334 - OpenAI Responses streaming tool-call transcript
• docs(xlsx): add a sanitized Excel oracle summary sample #383 - sanitized Excel oracle summary sample

Best first code and test tasks

Use these if you want one focused test and a small diff:

• test(headless): cover display-value readback after JSON restore #360 - display-value readback after JSON restore
• test(headless): cover range readback after an input edit #361 - range readback after an input edit
• test(examples): guard the headless README command index against missing scripts #362 - guard the headless README command index against missing scripts
• test(examples): add invalid-request proof to the HTTP JSON summary smoke #363 - invalid-request proof for the HTTP JSON summary smoke
• test(headless): cover changed named expressions after WorkPaper restore #366 - changed named expressions after WorkPaper restore
• test(headless): cover dense sheet range read with sparse values #367 - dense sheet range read with sparse values
• test(headless): cover two-column formula tiling in fill ranges #368 - two-column formula tiling in fill ranges
• test(headless): cover tab-indented formula prefix detection #369 - tab-indented formula prefix detection
• test(examples): add deterministic markdown-report output test #371 - deterministic markdown-report output test

Best first service and example docs

Use these if you want to add a copy-pasteable WorkPaper recipe:

• docs(examples): add Express WorkPaper route smoke #273 - Express WorkPaper route smoke
• docs(examples): add tRPC WorkPaper procedure smoke #300 - tRPC WorkPaper procedure smoke
• docs(examples): add Koa WorkPaper route smoke #330 - Koa WorkPaper route smoke
• docs(examples): add Elysia WorkPaper route smoke #331 - Elysia WorkPaper route smoke
• docs(examples): add Astro endpoint WorkPaper smoke #337 - Astro endpoint WorkPaper smoke
• docs(examples): add TanStack Start server function WorkPaper smoke #338 - TanStack Start server function WorkPaper smoke
• docs(examples): add Qwik City WorkPaper route smoke #339 - Qwik City WorkPaper route smoke
• docs(examples): add Next.js Server Action validation-error smoke #342 - Next.js Server Action validation-error smoke
• docs(examples): add Fresh WorkPaper route smoke #381 - Fresh WorkPaper route smoke

Agent, MCP, and evaluation tasks

Use these if you care about coding-agent workflows, MCP setup, or spreadsheet accuracy evaluation:

• docs(mcp): add MCP Inspector smoke-test transcript for the WorkPaper server #285 - MCP Inspector smoke-test transcript
• docs(mcp): add official registry API verification snippet #287 - official registry API verification snippet
• docs(mcp): add Claude Code setup for WorkPaper #289 - Claude Code setup for WorkPaper
• docs(mcp): add Codex MCP config for WorkPaper #290 - Codex MCP config for WorkPaper
• docs(agent): add OpenAI Responses streaming tool-call transcript #334 - OpenAI Responses streaming tool-call transcript
• docs(agent): add AI SDK onStepFinish WorkPaper transcript #358 - AI SDK onStepFinish WorkPaper transcript
• docs(agent): add Anthropic tool-use WorkPaper transcript #359 - Anthropic tool-use WorkPaper transcript
• docs(agent): add Firebase Genkit WorkPaper tool recipe #376 - Firebase Genkit WorkPaper tool recipe
• docs(xlsx): add a sanitized Excel oracle summary sample #383 - sanitized Excel oracle summary sample
• docs(mcp): refresh PulseMCP directory indexing proof #384 - PulseMCP directory indexing proof
• docs(google-sheets): add a Sheets API handoff checklist #385 - Google Sheets API handoff checklist

Before opening a PR

A useful claim comment is:

I can take this.

Plan:

- files I expect to touch:
- validation command:
- question or assumption:

Then keep the PR focused:

• Use TypeScript examples, not pseudocode.
• Prefer public @bilig/headless exports.
• Include the command you ran and the expected readback or output.
• Link the issue in the PR body.
• Open the pull request on GitHub.

Contributor docs:

• https://github.com/proompteng/bilig/blob/main/CONTRIBUTING.md
• https://github.com/proompteng/bilig/blob/main/docs/new-contributor-guide.md
• https://github.com/proompteng/bilig/blob/main/docs/starter-issues.md
• https://proompteng.github.io/bilig/starter-issues.html

[gregkonush]

Closing this internal docs-backlog placeholder to keep the public issue queue focused and credible. The current contributor path is intentionally capped at the scoped starter list in https://github.com/proompteng/bilig/blob/main/docs/starter-issues.md; broader docs ideas should be reopened only when they have an owner, exact files, and validation commands.