Sundae Bar Logo
Past Challenge
Internal Newsletter Curation & Briefing Skill
SBC14

Internal Newsletter Curation & Briefing Skill

Overview

Given a heterogeneous pool of raw candidate updates — Slack excerpts, status bullets, announcement drafts, team wins — plus an audience profile, a length target, and optional edition context, the agent must produce a structured content brief that tells a newsletter writer exactly what to include, where to put it, how much space to give it, and why each decision was made.

Your submission is a single SKILL.md file that teaches a fixed base agent to perform this curation and briefing task reliably across any audience, any update pool, and any length target it has not seen before.

Strong skills teach a reasoning procedure: how to normalise and deduplicate the incoming pool, how to score items on impact and audience relevance, how to let those two axes interact rather than treating them independently, how to allocate word-count budget proportionally, and how to flip priorities when the audience profile changes. Weak skills provide fill-in-the-blank templates or coach the agent to always include as many items as possible under a 'comprehensiveness' rationale — both break the moment surface details rotate.

The brief the agent produces must be organised into labeled sections (Lead Story, Need-to-Know, Quick Updates, FYI/Around the Team, Excluded). Each included item carries a title, source, Impact score, Audience-Relevance score, rationale, writer angle, and word-count budget. Excluded items carry a one-line exclusion reason pointing at audience mismatch or impact ceiling — not a restatement of the score.

A skill built to memorise specific scenarios will not hold up as challenges evolve.

You submit a SKILL.md file. A fixed base agent loads it and answers all 10 scenarios with it, then returns a scored breakdown. The submission that holds the top of the leaderboard accrues on-chain emissions for as long as it stays first.

We are looking for skills that demonstrate good design and help the agent perform like a capable practitioner in this domain.

Requirements (10 Tests)

Your skill is tested across four capability types, with 10 scenarios in total. Every submission is evaluated against all 10 scenarios — there is no sampling, so consistent quality across every capability type counts.

CategoryOverviewTest cases
Type A — Audience-Driven Item ScoringThis type tests whether the agent can score each candidate item on both an Impact axis and an Audience-Relevance axis, and then reason from the interaction of those two scores rather than from either alone. The skill must teach the agent to recognise that a high-impact item pitched to the wrong audience and a perfectly relevant item with a low impact ceiling both fail the brief — and to surface that reasoning transparently in the brief output.1
Type B — Persuasive Inclusion RationalesThis type tests whether the agent can write rationales for included items that are specific, audience-grounded, and genuinely persuasive — the kind a skeptical Chief of Staff would accept without pushback. The skill must teach the agent to move beyond generic justifications and articulate exactly why this item earns space for this audience at this moment, connecting it to concrete stakes or decisions the audience faces.1
Type C — Audience-Driven ReprioritisationThis type tests whether the agent radically reshuffles section assignments when the same candidate pool is evaluated against a different audience profile. The skill must teach the agent that no item has a fixed rank — its placement is entirely a function of who is reading — and that the procedure for assigning sections must be rerun from scratch when the audience changes.1
Type D — Risk Signal ElevationThis type tests whether the agent can identify risk signals, regulatory flags, or early-warning items in the candidate pool and elevate them appropriately for an audience that acts on such signals — even when those items lack the surface drama of a milestone announcement. The skill must teach the agent to write rationales for these elevations that make the downstream consequence concrete and audience-specific rather than generically cautionary.1

Surface details — names, organisations, item counts, word targets, and specific update content — rotate across variants while the underlying curation and briefing capabilities being tested remain the same.

Datasets

Available Datasets

When you submit, we run your SKILL.md through a standardised base agent provided by sundae_bar. The base agent is the same for every submission and is not supplied by you. It receives the challenge's test scenarios one at a time, calls the load_skill tool to read your skill, and follows it to answer each scenario.

Same base agent, same tests, only the submitted skill changes. Scores therefore reflect the quality of your skill, not agent configuration: the agent's answers are only as good as the procedure your skill teaches.

Graders then score each answer for output quality, correct skill use, and gaming behavior, and the per-scenario scores are averaged into your challenge score.

Important: the base-agent harness is fixed at evaluation time and its available tools are load_skill. You submit only a SKILL.md; you cannot supply your own agent, add tools, or alter the base agent's configuration.

Tip: write your skill as though you are onboarding a capable new hire who needs a procedure, not scripting a chatbot.

Evaluation Criteria

Each scenario is graded by an LLM judge. Final scores combine two graders and a set of deterministic penalties.

CriteriaWeight

Skill use

An LLM judge evaluates whether the agent's output actually followed the procedures your skill teaches, rather than freelancing or pretending to use the skill.

50 %

Scenario quality

An LLM judge evaluates how well each output meets the per-type quality bar, independent of whether your skill was followed. A reference good output anchors the judge's sense of the quality bar.

50 %

Scenarios are scored on the quality of the structured brief the agent produces; deterministic penalties apply when the agent makes moves that materially damage output quality — such as including items without rationales, producing newsletter prose instead of a brief, padding sections rather than cutting, or failing to ask for missing inputs before scoring. Individual scenario scores roll up to capability-type averages, which combine into the final challenge score shown on the leaderboard.

Submission

How to submit

1
Login or create a sundaebar.ai account

Create an account and add your Bittensor wallet ID to your profile to earn alpha rewards in the challenge.

2
Review the requirements

Review the reference skills (good, mediocre, gaming) to understand the quality bar and calibration points.

3
Build your skill

Draft your SKILL.md. Focus on teaching procedure rather than providing templates.

4
Test locally

Validate locally using the published dataset variants to sanity-check output quality.

5
Submit via the platform

Submit through the sundae_bar challenge portal by uploading your SKILL.md file.

6
Iterate

Review your scored breakdown and resubmit.

Constraints

Submission format: a single SKILL.md file with YAML frontmatter (name, description) and Markdown body.

Base agent: fixed. Provided by sundae_bar.

Runtime: a fixed base-agent harness loads your SKILL.md and answers all 10 scenarios per submission with it.

Tests per submission: 10 (the full scenario set — no sampling).

Submission limits: 5 submissions per day. A rolling challenge limit also applies.

Rewards

Emissions accrue on-chain to whichever submission currently holds the top of the leaderboard. There is no fixed per-submission payout: the current leader earns for as long as it stays first.

Time at the top takes all. A strong skill submitted early starts accruing immediately and keeps earning until another submission displaces it. Once you are displaced, accrual stops until you reclaim the top.

Rankings shift as others iterate. Get to the top early, then keep refining your skill to defend the position.

Your SKILL.md is scanned at submission time by a skill integrity gate. The gate checks that your skill teaches procedure rather than encoding scenario-specific content from our evaluation pool.

A skill that fails the integrity gate scores 0 for the submission. The gate runs deterministically and reports its finding as part of your results. This is how the evaluation framework stays credible: a high score should reflect a skill that genuinely teaches procedure, not one that memorised our test set.