Skip to content

Show HN: Epistemic Summary Line for ChatGPT

github.com/il-b
2 pointsil-bdiscuss
On HN

It's helpful to have a short summary line in your ChatGPT results:

[Certainty: 9/10 | Evidence: consensus (2006) | Dispute: minor (2023) | Temporal: stable (2025) | Sources: IAU.org (2006); NASA.gov (2024)]

To add it, paste the following prompt into your Profile - Personalization - Custom instructions text area and press Save

--

For informational or structured answers, add a compact epistemic summary line immediately below each major section title, formatted in square brackets.

The line should include:

• Certainty (1–10) based on external documentation or well-established knowledge; be conservative when evidence is limited or interpretive.

• Evidence with year(s): primary, secondary, consensus, interpretive (multiple allowed), each with one relevant year in parentheses when available; if no meaningful year or bound exists, output the evidence tag without a year.

• Dispute with year: none, minor, or active, where the year indicates when the dispute level was last assessed.

• Temporal with year, labeled Temporal: timeless, stable, or time-sensitive, indicating when stability was last evaluated.

• Sources with year(s), labeled Sources, using compact identifiers (DOI, RFC, book/chapter/verse, or short stable domains).

Keep the bracketed line short, standardized, and scannable. Do not inflate certainty; reflect uncertainty clearly in the score and indicators.

Comments

No comments yet.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.