agent-friction-ledger
Skill16 starsChanged 5 months agoMITView on GitHub
Capture recurring agent-development friction as a local, privacy-safe Markdown report. Use only when the user explicitly requests a friction review of blockers, misleading errors, missing documentation, incompatible tools, or manual workarounds that should improve a doc, tool, skill, process, or evaluation.
In this repository94 other files it gives its agents.
- 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
- filesystem-context-disciplineSkill
- 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
6 rules. About 0 are common, 5 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
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
Do not run merely because a task had normal shell output, a single failed command, or a clean completion.
- Distinctive: line 32
Do not record every tool call, full conversation, raw logs, private URLs, credentials, personal data, client data, or unrelated …
- Very specific: line 40
Write a local report to an ignored or private location only.
- Distinctive: line 71
Never write to a tracked repo path by default.
- Distinctive: line 73
Prefer one concise report over a growing per-turn buffer.
- Distinctive: line 74
If the task contains regulated, personal, client, deal, financial, or credential data, record the abstract failure mode only.
≈ “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
Agent Friction Ledger
Turn repeatable agent friction into a small, actionable local artifact. This skill is opt-in and never sends session data outside the workspace.
Trigger
Use only when:
- the user explicitly asks for a friction report or review.
Do not run merely because a task had normal shell output, a single failed command, or a clean completion.
Capture Rules
Record only friction that is actionable and evidenced by the current task:
- repeated blocker or retry loop;
- misleading, incomplete, or contradictory error;
- missing, stale, or hard-to-find documentation;
- tool, provider, environment, or version incompatibility;
- manual workaround that should become a durable improvement.
Do not record every tool call, full conversation, raw logs, private URLs, credentials, personal data, client data, or unrelated working-tree changes.
Workflow
- Name the task boundary and identify only the material friction events.
- For each event, record:
symptom -> attempted -> root cause -> resolution -> durable fix. - Mark confidence: confirmed, likely, or unresolved. Do not turn a guess into a root cause.
- Redact or omit sensitive content before writing.
- Write a local report to an ignored or private location only. Default to
.git/agent-friction/YYYY-MM-DD-<task>.mdunless the repo defines another approved ignored path. - Group action items under
docs,tooling,skill/process, andresearch. - Propose the smallest durable change, but do not apply it without normal task authorization.
Report Shape
# Agent Friction Report: <task>
## Scope
## Material Friction
### <short label>
- Symptom:
- Attempted:
- Root cause: <confirmed|likely|unresolved>
- Resolution:
- Durable fix:
## Action Items
### Docs
### Tooling
### Skill / Process
### Research
## Decision
- Keep local / propose follow-up / no durable changeGuardrails
- Local Markdown only. Never POST to a viewer, telemetry endpoint, or external API.
- Never write to a tracked repo path by default. If no ignored/private location exists, stop and ask before writing.
- A friction report is not permission to create a ticket, alter a skill, edit docs, or change a process.
- Prefer one concise report over a growing per-turn buffer.
- If the task contains regulated, personal, client, deal, financial, or credential data, record the abstract failure mode only.
Pair With
repo-debugging-ledgerfor hypothesis-driven technical diagnosis;durable-context-maintenancewhen an accepted fix belongs in repo context;token-usage-auditwhen the friction is context or tool-output waste;skillopt-evolve-skillsandbounded-evaluationwhen an accepted skill change needs validation.
Provenance
Adapted from aurorascharff/agent-friction-skill, narrowed to explicit, local-only, privacy-safe friction capture with proposal-gated follow-up.
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.

