idea-pipeline

${var} — Optional theme filter (e.g. "crypto", "AI agents", "consumer"). If empty, scans all ideas. A pick:<id|name> value (from the "build next?" force-reply — e.g. pick:2 or pick:Onchain reputation) instead marks that idea as chosen-to-build in the backlog and ends, skipping the audit — see step 0. offer:<owner/repo or issue-url> directly offers a confirmed owned target through the same Telegram reply boundary, also ending before the audit.

Today is ${today}. Read memory/MEMORY.md before starting. If soul/SOUL.md + soul/STYLE.md exist and are populated, read them to ground "operator fit" scoring; otherwise score on the idea's general buildability and timing alone.

Why this skill exists

idea-forge validate evaluates ideas. Nothing tracks execution. Backlogs of dozens of ideas accumulate — some validated, most unscreened — with zero visibility into which ones have been acted on vs which are rotting. This skill gives that view: pipeline size, execution rate, and the 3 ideas closest to being buildable right now.

Steps

0. Force-reply interception — pick:<idea> / offer:<target> (run FIRST, before anything else)

If ${var} starts with offer:, strip and trim the remainder. Accept only owner/repo or https://github.com/owner/repo/issues/N, normalize it to owner/repo, and require gh api "repos/$repo" --jq '.permissions.push // false' to return true. Invalid, inaccessible, or API-failed targets get a plain rejection notification and end without a force reply. For a confirmed target, execute this exact command with Bash. Describing or printing the command is not delivery:

./notify "Which owned repository or issue should Aeon Engineer use? Reply with ${target}." \
  --force-reply --placeholder "${target}" \
  --context "dev-loop::ship"

After the command exits zero, require at least one non-empty JSON payload under ${AEON_PENDING_DIR}/notify-queue/ and that payload's .reply_markup.force_reply == true. A queued payload alone isn't enough — notify.sh queues one even when the inbound Messages workflow is disabled, but in that case sends the prompt as plain text (reply_markup:null) with no reply routing, so the operator's answer would never come back to this skill. If the command fails, the payload is absent, or force_reply isn't true, end with FORCE_REPLY_MISSING: dev-loop::ship target=<target> and do not claim the prompt was offered. Only after both checks pass, log FORCE_REPLY_OFFERED: dev-loop::ship target=<target> under ### idea-pipeline, then end. This is an explicit operator-invoked producer path and still does not dispatch the chain until the operator replies.