More screenshots
Works with the AI tools you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+22 moreClaude CodeCursorCodex CLIMuseDots
OpenClaw
Manus
Grok
Grok Bot+18 agentsContext Rot Forensics
Separate position loss from fill saturation in failing long-context AI sessions.
$19.99
Context Rot Forensics
Example session with this skill installed
Diagnose this 40-turn coding session. The database schema appeared near the beginning, the session is now about 75% full, and later output stopped respecting that schema. Evidence E3 confirms the schema was still present but missed. Distinguish position effects from fill saturation, propose only justified repair moves, and define an A/B rerun with pass criteria.
- Read your context and instructions
- Compiled the context rot forensics
Context Rot Diagnosis
Verdict: FILL SATURATION
Estimated fill: 75%
Missed item: db_schema
Observed zone: HEAD
Depth: approximately 15% from the start
Evidence cited: E3
The critical schema was not buried in the BODY zone; it remained in the HEAD. That weakens a pure position-rot explanation. The stronger supported diagnosis is saturation at high fill, with the missed schema becoming harder to use reliably as the session accumulated turns. This is an evidence-bounded classification, not a claim about hidden model internals.
Repair moves
O5 — Applied inference: create a fresh state summary that pulls the schema and current decisions into a compact TAIL position.
2.
O7 — Applied inference: reset the session with a structured handoff at roughly 75% fill instead of continuing to accumulate context.
A/B placement test
- A: rerun the next task with the original 40-turn context unchanged.
- B: rerun the same task with the schema and live decisions in a structured TAIL handoff.
- Keep model, prompt, temperature, and evaluation criteria constant.
- Pass condition: B follows every required schema field while A reproduces at least one omission.
If both variants pass or both fail, the repair is not proven and the verdict remains unresolved for that task.
Connects securely to your tools. The creator never sees your data.
What you get
About this skill
Diagnose whether a long-context failure is position rot, fill saturation, a mixed case, or not context rot. Context Rot Forensics maps critical information across HEAD, BODY, and TAIL zones, estimates context fill, cites the evidence behind the missed item, distinguishes observation from inference, and proposes a layout repair only when the diagnosis supports one.
The deliverable includes a scope-limited verdict, missed-item map, evidence ledger, repair moves, and an A/B placement test with exact pass criteria. The package includes the full SKILL.md workflow, CROT protocol, structured sample, deterministic context-map checker, tests, and a six-page PDF buyer guide.
It does not increase a model's context window, guarantee recall, summarize documents, or claim a repair worked without the supplied A/B result.
How to install
Works the same in every agent - Claude, Cursor, Codex, Copilot and 20+ more.
- 1
Download the ZIP
Free skills download straight away. Paid skills unlock right after purchase.
- 2
Unzip into your skills folder
Every agent reads skills from one folder on your machine. Drop the unzipped folder in there.
- 3
Ask your agent to use it
Restart the agent if it was already running. It picks the skill up automatically - no config needed.
Skills folder by agent
Click the path to copy it. Create the folder if it does not exist yet.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
30-day refund
Not a fit? Get your money back
Trust & safety
Security scanned
Verified clean 1 month ago
- Passed all security checks, Safe to install