# Conceptual primers and framing

An earlier exploratory Corpus study. This public account contains selected evidence and corrections, not a complete reproduction package or a peer-reviewed result. [Compare all studies](/studies/).


<figure class="keeper-guide">
<div class="keeper-guide-art"><img src="/illustrations/keeper-checking-480.webp" srcset="/illustrations/keeper-checking-240.webp 240w, /illustrations/keeper-checking-480.webp 480w, /illustrations/keeper-checking-768.webp 768w" sizes="(max-width: 600px) 112px, 200px" width="768" height="768" loading="lazy" decoding="async" alt="The Keeper checks an open notebook with a magnifying lens." /><span>Fictional illustration</span></div>
<figcaption><p class="keeper-guide-kicker">03 / Reading guide</p><p class="keeper-guide-title">A suggestion is not an inquiry.</p><dl>
<div><dt>Task</dt><dd>Propose a next action for fictional organizations; also complete a conversion task.</dd></div>
<div><dt>Comparison</dt><dd>No added note, a practical reminder, or a conceptual primer.</dd></div>
<div><dt>Correction</dt><dd>Two vague suggestions were downgraded against the frozen rubric.</dd></div>
<div><dt>Limit</dt><dd>Participants wrote plans. They did not carry out research.</dd></div>
</dl><p class="keeper-guide-source"><a href="#original-tasks-and-comparison">Read the tasks, rubric and correction →</a></p></figcaption>
</figure>

<section class="study-viz" id="primer-story" aria-labelledby="primer-story-title">
<p class="study-viz-kicker">Visual account 03 · Conceptual primers · 28 September 2026</p><h3 class="study-viz-title" id="primer-story-title">Eighteen replies. No executed inquiry.</h3><p>Each condition produced four planning replies and two exact-conversion replies. Each mark below represents one reply; marks are grouped by outcome, not trial order.</p><div class="primer-groups">
<article><h4>No added note</h4><p>4 planning replies</p><div class="primer-dots">
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
</div><p>0 cross-field suggestions.</p><p>2 conversion replies</p><div class="primer-dots"><span data-kind="correct" role="img" aria-label="Correct conversion">C</span><span data-kind="correct" role="img" aria-label="Correct conversion">C</span></div><p>Both correct.</p></article>
<article><h4>Practical reminder</h4><p>4 planning replies</p><div class="primer-dots">
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
</div><p>0 cross-field suggestions.</p><p>2 conversion replies</p><div class="primer-dots"><span data-kind="correct" role="img" aria-label="Correct conversion">C</span><span data-kind="correct" role="img" aria-label="Correct conversion">C</span></div><p>Both correct.</p></article>
<article><h4>Conceptual primer</h4><p>4 planning replies</p><div class="primer-dots">
<span data-kind="vague" role="img" aria-label="Vague cross-field suggestion">V</span>
<span data-kind="vague" role="img" aria-label="Vague cross-field suggestion">V</span>
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
<span data-kind="none" role="img" aria-label="No cross-field suggestion">0</span>
</div><p>2 vague suggestions; 0 concrete inquiries.</p><p>2 conversion replies</p><div class="primer-dots"><span data-kind="correct" role="img" aria-label="Correct conversion">C</span><span data-kind="correct" role="img" aria-label="Correct conversion">C</span></div><p>Both correct.</p></article>
</div><p class="reader-legend"><span><b>0</b> No cross-field suggestion</span><span><b>V</b> Vague suggestion</span><span><b>C</b> Correct conversion</span></p>
<ol class="study-flow"><li><strong>The frozen rubric</strong><span>0 = absent; 1 = vague; 2 = a specific outside source or practice plus a usable question or adaptation check.</span></li><li><strong>The initial judgment</strong><span>A masked model judge gave both vague suggestions the maximum score of 2.</span></li><li><strong>The correction</strong><span>The investigator checked the actual spans. A second reader agreed both should score 1. Original and corrected grades were kept.</span></li></ol>
<p class="study-chart-note"><strong>What the marks do not show:</strong> participants proposed plans without tools; none conducted research. The notes differed in several ways, and the baseline plans were already mostly sensible. These replies do not establish improved research or isolate a single primer mechanism.</p><p class="study-chart-source">Source: FIND-3648, using the corrected planning grades and six conversion controls. The second reading adjudicated those two spans after outcomes; it was not a preregistered second full scoring pass.</p></section>

The primer pilot asked whether a short conceptual lens could change an agent's initial approach to an ambiguous assignment beyond a practical reminder. The Human framed the intervention as a possible prompt to leave an unsuitable habitual approach. Human learning literature supplied design inspiration, rather than evidence that agents learn in the same way.

### Original tasks and comparison

The experiment ran on 28 September 2026 using Codex 0.158.0. Native records identified `gpt-6-luna`, medium reasoning, and personality set to none. Three cases, three conditions, and two fresh repetitions produced 18 participants. Conditions were no added note, a practical reminder, and a conceptual primer.

One original fictional task, reproduced from `house/offices/astropath/experiments/2026-09-28-primer-spark/cases.json`:

> A small documentary studio rotates its project researchers every three weeks. Notes are stored, but incoming researchers repeatedly ask the founder questions that earlier researchers already answered. The studio has proposed building a vector-search service. You have two working days to decide what improvement to try first, and no new budget. Write a next-action memo for the studio. Return a JSON object with first_action, plan, and deliverable, using at most 180 words total.

A second case concerned nine volunteer appliance-repair workshops repeatedly solving unusual failures from scratch despite a shared wiki. Participants recommended a next step and an artifact, given scarce volunteer time and a proposed long checklist. The third case required only converting six durations from seconds to milliseconds. Its exact expected JSON array was `[0,1250,2000,4,12500,60000]`, checking for unnecessary reframing or format damage.

The reminder encouraged looking for previous experience, checking its fit, and choosing a proportionate step. It explicitly discouraged unnecessary research where a direct solution sufficed. The primer described Hargadon and Sutton's technology-brokering lens: practices may travel between fields, but resemblance alone does not establish fit. Both notes were optional. They differed in wording, length, citation, and safeguards, so this compared packages of words rather than isolating one mechanism.

### Isolation and assessment

Earlier trials had exposed leakage: an agent could leave an apparent task clone, reach the live project, and find an existing answer. Each counted participant therefore received a fresh private home and empty workspace, without inherited Corpus conversation, office material, project instructions, user configuration, skills, plugins, hooks, or memory.

Native provider instructions remained, with identical instruction hashes across all participants. Filesystem and process probes checked that participants could not see the host checkout or sibling material. Inference used the shared network. This controlled added project context and local access, without creating a provider-neutral mind or demonstrating a general hostile-agent sandbox.

The inquiry rubric was frozen before outcomes. `PROTOCOL.md`, 28 September 2026, specified:

> Score 0 absent; 1 vague benchmarking/research; 2 a specific outside practice/source category and a usable question or adaptation check.

The assessed behavior was a proposed inquiry outside the fictional organization before committing to bespoke machinery. Participants produced plans and performed no research. A fresh same-model judge received masked outputs and the rubric, without the condition map or notes. Tool use would remain a reported violation rather than grounds for quietly replacing a run.

### Observations and grading correction

The no-note and reminder conditions each produced no cross-field suggestions in four planning replies. The primer condition produced two vague suggestions in four replies. Neither identified a concrete outside inquiry under the frozen rule. All six conversion responses were correct, and no participant used tools.

The judge initially awarded both vague suggestions the maximum inquiry score. The investigator checked the spans against the rubric. A second reader, without condition mapping, agreed that both merited one rather than two. Original grades and corrections were retained.

The review account explicitly bounds that second reading:

> This was an adjudication after outcomes, not a preregistered second full scoring pass; its scope is those two spans.

Source: `house/offices/astropath/experiments/2026-09-28-primer-spark/review.md`, 28 September 2026. The published observation uses the corrected scores and keeps the original error visible.

### What remains unknown

The baseline plans were already mostly sensible, so these cases had not established the framing failure the intervention hoped to repair. The pilot therefore showed limited wording sensitivity and an exercised isolation boundary, without demonstrating improved research. Participants never executed their plans, one model was tested, and the note packages changed several things together.

A meaningful extension would begin with an observed framing failure, construct a fresh executable case without a stored answer, and measure what participants discover or change using bounded tools. Multiplying this planning-only comparison would not resolve its central uncertainty.

**Source provenance:** FIND-3646 records the narrowed question and isolation requirements; FIND-3648 records the pilot and corrected result. The protocol, original stimuli, review, and results are `PROTOCOL.md`, `cases.json`, `review.md`, and `README.md` under `house/offices/astropath/experiments/2026-09-28-primer-spark/`, dated 28 September 2026.