Obliviate Fossilium

---
name: obliviate-fossilium
description: Put memory files (CLAUDE.md, AGENTS.md) on a diet — find fossils (migration-relative phrasing, "this used to say X" narratives, dated retractions living in the hot path), typed facts that drift (hand-written counts, sizes, versions), and history that belongs in an archive. The 5-family reads the whole file every session; every stale or historical sentence is context spent steering nothing. Use for "my CLAUDE.md is huge", "Claude cites outdated facts", "clean up my memory file", or as part of /opus-pocus. Prescribes current-fact rewrites, derivation commands, and an archive file. NOT for deleting live rules — Avada Verbosa owns ablation of things that might still work.
---

# ✨ Obliviate Fossilium

> *"Your memory file remembers things that stopped being true in July."*

Memory files accrete like sediment: every correction keeps the old wording as a warning,
every migration leaves "note: this changed" markers, every count gets hand-typed and never
updated. The model reads all of it, every session, and weighs the fossil as heavily as the
living rule.

## What to scan

The root memory file, every nested one, whatever the repo names them (`CLAUDE.md`, `AGENTS.md`),
user-global memory files with the user's consent, and any doc the instruction layer tells agents
to read routinely.

## Anti-patterns to find

1. **Migration-relative phrasing.** "This used to say X until <date>", "note this changed",
   "the old rule was…". Useful for the week after the change; a fossil forever after. The
   current fact is the only fact the model needs.
2. **Retraction narratives in the hot path.** Paragraph-length stories of how a claim was
   wrong and got corrected. The correction earns its context; the story doesn't.
3. **Typed facts that drift.** Hand-written counts, sizes, version numbers, dates ("the suite
   has 5,738 tests") that go stale silently. Symptom: two different numbers for the same fact
   in one file.
4. **History as instruction.** Vendor-adoption sagas, decision back-stories, who-ruled-what
   timelines living inline where a decision-record pointer would do.
5. **Restated rules.** The same rule paraphrased in three sections because each author feared
   the reader would miss it (also feeds Contradicto Revelio when the copies drift apart).
6. **Panic-marker sediment.** ⚠ markers on resolved history (→ Quietus Capslockius grades the
   volume; this pass relocates the content).
7. **Sheer size.** Official guidance targets lean memory files (on the order of a couple
   hundred lines); multi-thousand-line roots are paying context tax every single session.

## Fix prescriptions

| Finding class | Fix |
|---|---|
| Migration-relative phrasing | State only the current fact; keep the pointer to the decision record if one exists |
| Retraction narratives | Move the narrative VERBATIM to an archive file (`docs/history.md` or the repo's knowledge dir); the hot path keeps one sentence |
| Typed facts | Replace with the derivation: the command that computes the fact (`ls … \| wc -l`), stated as "derive it, don't quote it" |
| History as instruction | One-line pointer to the ADR/knowledge file; the saga lives there |
| Restated rules | Single-source: the rule once, in its home section; pointers elsewhere |
| Marker sediment | Markers only on live hazards; archive the rest with the narratives |
| Size | After the above, re-measure. If still oversized, hand candidates to Accio Rulebook (path-scoped extraction) and Avada Verbosa (ablation) |

## The archive is not a landfill

The extracted history must stay findable: one archive file, dated sections, indexed from
wherever the repo indexes knowledge. Deleting history outright destroys the evidence trail
that keeps decisions from being re-litigated — move it, don't burn it.

## Bounds

- **If `legilimens-anthropica` ran this session, prefer its measured numbers over the ones in this file.** This file is a snapshot; that pass is the current reading.
- Scope: memory files — root, nested, and user-global only with the user's consent. Not application source, not docs the instruction layer merely links.
- Cap: 15 findings, ranked by bytes recovered per unit of risk. Class counts carry the rest.
- Done when each memory file has been read once and its fossils classified.
- Quote once, at read time; mark anything you cannot re-confirm `UNVERIFIED` rather than re-reading to check yourself.
- See `skills/opus-pocus/references/casting-safety.md` for the full rules.

## Report format

Per finding: `file:line` · class (1–7) · quoted text · destination (rewrite in place / archive /
derive). Lead with the before/after: bytes now, projected bytes after, fossil count by class.
`SCANNED NOTHING` if no memory files exist, grading `N/A`.