docs: point primary agent docs to the issue board

AGENTS.md, QUICKSTART.md, CONTRIBUTING.md, and llms.txt now surface the
agent-task issue board as the fastest discovery + claim path, with a
good-first-agent-task entry point for newcomers.

Author: cschanhniem

AGENTS.md

@@ -6,6 +6,10 @@ This repository is a GitHub-native protocol for verified contributions to neglec
 
 If a change does not survive `pnpm validate`, replication, and human review, it is not knowledge.
 
+## Fastest Way In: The Issue Board
+
+Scoped tasks are published as GitHub Issues so you can discover and claim work without reading the whole repo first. Browse [`type:agent-task`](https://github.com/Open-Problem-Lab/open-problem-lab/issues?q=is%3Aissue+is%3Aopen+label%3Atype%3Aagent-task), start with [`good-first-agent-task`](https://github.com/Open-Problem-Lab/open-problem-lab/issues?q=is%3Aissue+is%3Aopen+label%3Agood-first-agent-task), comment to claim (one agent per task), then submit via the Agent Submission template or a PR. See [`docs/AGENT-ISSUES.md`](docs/AGENT-ISSUES.md) for how the board works and how it is generated from `tasks-available.json`.
+
 ## Why Strong Contributors Should Work Here
 
 Most public AI output is unverifiable, unranked, and unowned. This repository inverts that:

CONTRIBUTING.md

@@ -8,6 +8,8 @@ Read [VISION.md](VISION.md) to understand why verification is the product here,
 
 Browse [problem-packs/](problem-packs/) to find a problem that matches your expertise. Each pack contains a `task-map.md` and `tasks.json` with specific scoped work — these are your starting points.
 
+The quickest way to find claimable work is the [agent-task issue board](https://github.com/Open-Problem-Lab/open-problem-lab/issues?q=is%3Aissue+is%3Aopen+label%3Atype%3Aagent-task). Comment on an issue to claim it (one contributor per task), then submit via the matching issue form or a PR. See [docs/AGENT-ISSUES.md](docs/AGENT-ISSUES.md).
+
 ## Contribution Types
 
 | Type     | What it is                                             | Issue form      |

QUICKSTART.md

@@ -12,6 +12,8 @@ Read these in order — they take about 10 minutes together and prevent the most
 
 ## Step 1: Pick a Problem
 
+The fastest path of all: browse the [open agent-task issues](https://github.com/Open-Problem-Lab/open-problem-lab/issues?q=is%3Aissue+is%3Aopen+label%3Atype%3Aagent-task) and claim one by commenting. New contributors should filter for [`good-first-agent-task`](https://github.com/Open-Problem-Lab/open-problem-lab/issues?q=is%3Aissue+is%3Aopen+label%3Agood-first-agent-task). Each issue is self-contained — see [docs/AGENT-ISSUES.md](docs/AGENT-ISSUES.md).
+
 If you want the shortest path instead of browsing manually, open [agent-radar.json](agent-radar.json) first. It ranks the best first moves, the packs that unlock the most follow-on work, and the reviewer bottlenecks likely to matter after the first task lands.
 
 Browse [problem-packs/](problem-packs/) and find a domain that matches your expertise or curiosity.

llms.txt

@@ -16,6 +16,10 @@ The bottleneck in addressing global problems is not generating candidate answers
 
 [SKILL.md](SKILL.md) — street-smart contribution patterns: narrow done conditions, kill conditions, decay handling
 
+[Open agent-task issues](https://github.com/Open-Problem-Lab/open-problem-lab/issues?q=is%3Aissue+is%3Aopen+label%3Atype%3Aagent-task) — claimable scoped tasks; comment to claim, start with `good-first-agent-task`
+
+[docs/AGENT-ISSUES.md](docs/AGENT-ISSUES.md) — how the GitHub Issue task board works and how it is generated
+
 [tasks-available.json](tasks-available.json) — machine-readable index of every scoped task ready to pick up right now
 
 [docs/AGENT-FAQ.md](docs/AGENT-FAQ.md) — common rejection patterns and how to recover