filesystem-context-discipline
Skill16 starsChanged 5 months agoMITView on GitHub
Use files as durable context without polluting the live prompt. Apply when a task needs scratchpads, plan persistence, retained outputs, proposal folders, handoffs, or repo-local context that agents can discover and audit later.
In this repository94 other files it gives its agents.
- agent-friction-ledgerSkill
- agent-learning-layer-triageSkill
- agent-operations-contractSkill
- agents-best-practicesSkill
- agent-tool-contract-reviewSkill
- anti-slop-reviewSkill
- architecture-deepening-reviewSkill
- assumption-excavationSkill
- authorized-security-routerSkill
- book-to-skillSkill
- bounded-evaluationSkill
- bounded-growth-loopSkill
- bounded-orchestration-contractSkill
- brief-first-executionSkill
- browser-verificationSkill
- bug-evidence-protocolSkill
- codex-review-loopSkill
- compaction-survivalSkill
- complexity-optimizerSkill
- confidence-fragility-reviewSkill
- context-degradation-reviewSkill
- corporate-surfaceSkill
- delivery-baseline-auditSkill
- delivery-preflight-gateSkill
- design-critique-polishSkill
- designmd-brand-kitSkill
- design-register-bootstrapSkill
- diagnoseSkill
- doc-grounded-grillingSkill
- doc-to-lora-evaluatorSkill
- durable-context-maintenanceSkill
- dynamic-workflow-packetsSkill
- editorial-surfaceSkill
- evidence-ledger-researchSkill
- fluid-interaction-reviewSkill
- frontend-product-builderSkill
- frontend-taste-layerSkill
- git-native-context-contractSkill
- goal-loop-designerSkill
- graph-guided-code-readingSkill
- handoffSkill
- happy-handoffSkill
- html-brief-artifactSkill
- html-diagram-artifactSkill
- hypothesis-diversificationSkill
- idea-to-ship-gatesSkill
- knowledge-base-enrichmentSkill
- lean-context-layoutSkill
- local-inference-tuningSkill
- loop-hardening-contractSkill
- loopops-protocolSkill
- loop-readiness-reviewSkill
- lottie-motion-builderSkill
- minimal-diff-builderSkill
- motion-review-gateSkill
- overengineering-reviewSkill
- personal-workspace-routerSkill
- phase-spec-executionSkill
- pipeline-readiness-gateSkill
- plan-to-issuesSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
10 rules. About 0 are common, 9 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈9
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 12
Use the filesystem to hold durable state, but keep it structured, discoverable, and reviewable.
- Distinctive: line 34
Put the artifact in the narrowest sensible location:
- Distinctive: line 39
Add enough metadata for later discovery: task, source, status, owner, date, and apply state when relevant.
- Distinctive: line 40
Keep scratchpads separate from durable docs.
- Distinctive: line 42
Remove or ignore transient files before commit unless they are intentional artifacts.
- Distinctive: line 57
Do not make every thought durable.
- Very specific: line 58
Do not mix speculative scratchpad text into README, AGENTS.md, or SKILL.md.
- Distinctive: line 59
Do not treat retained output as applied work.
- Distinctive: line 60
Do not use hidden files as the only record of an important decision.
- Distinctive: line 61
Do not write memory unless the user explicitly asked for memory persistence.
≈ “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
Filesystem Context Discipline
Use the filesystem to hold durable state, but keep it structured, discoverable, and reviewable.
Use When
- a task is too long for one prompt or needs resumable state;
- generated output should be retained before apply;
- a plan, checklist, evidence ledger, or handoff must survive compaction;
- multiple agents or sessions need a shared artifact;
- repo-local context should be discoverable without always loading it.
Context Types
scratchpad: temporary reasoning support, not authoritative.plan: current intended steps, assumptions, and done criteria.evidence: commands, test results, source links, or reviewed facts.retained_output: proposal artifacts kept separate from the target workspace.handoff: compact continuation state for another session or agent.durable_doc: stable repo guidance that future agents should read on demand.
Workflow
- Choose the file type before writing.
- Put the artifact in the narrowest sensible location:
- repo docs for durable guidance;
- task/proposal folder for retained output;
- benchmark or run folder for validation evidence;
- memory only when the user explicitly asks to preserve a cross-session lesson.
- Add enough metadata for later discovery: task, source, status, owner, date, and apply state when relevant.
- Keep scratchpads separate from durable docs.
- Link retained outputs to their inspection and settlement decision.
- Remove or ignore transient files before commit unless they are intentional artifacts.
Output Shape
Use:
- artifact type;
- path;
- authority level:
scratch,proposal,evidence, ordurable; - discovery cue;
- expiry or update trigger;
- related verification or settlement step.
Guardrails
- Do not make every thought durable.
- Do not mix speculative scratchpad text into README, AGENTS.md, or SKILL.md.
- Do not treat retained output as applied work.
- Do not use hidden files as the only record of an important decision.
- Do not write memory unless the user explicitly asked for memory persistence.
Provenance
Adapted from filesystem-context patterns in muratcankoylan/Agent-Skills-for-Context-Engineering, aligned with ABVX proposal-first and durable-context workflows.
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.

