session-story
Skill0 starsChanged 54 days agoMITView on GitHub
Extract a build session's story (narrative, problem-solving, wrong turns, the user's judgment calls, receipts) into blog-post source material at stories/. Use at the end of any substantive build session, when the user says "capture this session" or "story of what we did", or when mining an old transcript for a post. Raw material, never a drafted post.
2 other projects use a version of this.
Adopted by
Most recently taken up by
- nalaj/session-storytheir own version, first seen Oct 17, 2026
- heygen-com/hyperframes-community-skillstheir own version, first seen Aug 25, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- heygen-com/hyperframes-community-skills★ 155same name
- nalaj/session-story★ 22same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository5 other files it gives its agents.
- gif-loop-captureSkill
- release-gateSkill
- self-critiqueSkill
- ship-learn-nextSkill
- ui-auditSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
Numbers carry sources: Commit hashes, file sizes, token counts, wall-clock.
- Very specific: line 24
Commit the story file on its own, if the repo uses git: process: session story <slug>.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Session story extraction
Turn the session that just happened into durable source material for a blog post. This is UPSTREAM of drafting: a story file holds the narrative and the judgment, so a post can be written weeks later by a session that was not there.
FORMAT.md in this directory is the output format. Match its altitude: beats and receipts, not a summary.
Procedure
- Score first, write second: draft the full story into a scratch file, then gate it before it lands. It passes only if it has at least one wrong turn, at least one verbatim user quote, and a named invariant. A draft missing any of the three is low signal; stop at the source, do not file it.
- Output path:
stories/YYYY-MM-DD-<slug>.md(repo root). One file per story; a session with two unrelated arcs gets two files. - Fill every FORMAT.md section from the session in context. Running after the fact, mine the transcript or the git log for the receipts first, then write.
- Hard rules:
- the user's words verbatim. Their reactions, corrections, and one-liners are the judgment payload (a one-line reaction is worth more than any summary of it). Quote, don't polish.
- Failures stay in. The wrong turn, the probe that lied, the cache that bit: the debugging IS the story. A story with no friction is a press release, bounce it.
- Numbers carry sources. Commit hashes, file sizes, token counts, wall-clock. Unverifiable claims get cut, not hedged.
- Voice-neutral. This is source material; the user's voice gets applied at drafting time (a later drafting pass). No em dashes even here.
- The invariant is required. One to three sentences on what transfers beyond this session. If you cannot name one, say so explicitly; that is itself information.
- Append one line to
stories/INDEX.md:- [date] <slug>: <one-line hook> (status: raw). - Commit the story file on its own, if the repo uses git:
process: session story <slug>. Never sweep unrelated staged changes into it.
Relationship to the judgment-extraction loop
The format encodes a compressed version of the user's recursive judgment extraction (experience -> externalize -> tension -> invariant -> compress -> codify): Beats = experience externalized, Taste moments = the judgment located, The invariant = the compression, this skill = the codify step. The full ChatGPT-authored articulation is inspo, not canon; do not import its prose into stories.
Known gotchas
- Receipts survive a session; the hook, the wrong turns, the invariant, and the leftovers do not. Write them BEFORE clearing context, or they are gone.
- Prompting a model with texture targets (sentence-length mixes) produces a metronome; keep the prompt spec-only and edit a flat draft by hand.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.