agentleFS
gh-xjgh-xj/agent-repo-kit/skills/slock-epic-protocol/SKILL.md

slock-epic-protocol

Skill0 starsChanged 6 months agoMITView on GitHub

Use when designing, bootstrapping, auditing, repairing, or evolving a Slock-agent-to-epic-repo protocol: Slock agents mapped by slock_agent_registry to concrete leaf repos, generated pointer-only MEMORY.md files, generated repo symlinks, per-leaf tracked .work stores, repo-local skill mirrors, and cross-repo handoffs. Do not use for ordinary Slock chat/task commands or generic repo conventions without Slock agent mapping.

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 0 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Default work model is per_leaf_tracked: every leaf owns its own .work/.

    line 60
  2. Very specific:

    Use references/protocol.md to choose surfaces, ownership boundaries, work model, and runtime skill discovery.

    line 71
  3. Very specific:

    Use references/bootstrap.md to add descriptor fields, leaf domain claims, Taskfile commands, target-owned install/check scripts, and …

    line 77
  4. Very specific:

    Use references/audit-migration.md to compare descriptor, leaf repos, Slock agent dirs, memory files, notes, and skill roots.

    line 83
  5. Very specific:

    Use references/verification.md.

    line 110
  6. Very specific:

    If the task changes this skill's behavior, update MAINTENANCE.md in the same change.

    line 118

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Slock Epic Protocol

Maintain the bridge between Slock runtime agents and a federated epic repo. The protocol is repo-owned; Slock is the wake-up, social, and task surface.

Use This For

  • Creating or changing slock_agent_registry.
  • Mapping one Slock agent to one concrete leaf repo.
  • Installing or checking generated Slock repo symlinks (always) and epic symlinks (when epic_symlink_name is set on the registry — see references/protocol.md).
  • Keeping Slock MEMORY.md pointer-only.
  • Migrating durable Slock notes into repo docs, skills, or .work.
  • Adding repo-local Claude/Codex skill discovery for mapped agents.
  • Designing cross-repo handoffs where target repo inbox/work is canonical.

Do Not Use For

  • Sending Slock messages, claiming Slock tasks, or reading Slock history for an ordinary work request.
  • Generic epic-wrapper setup with no Slock agent map.
  • Generic skill authoring. Use skill-builder.
  • Generic .work/ operation. Use work-cli.
  • General repo convention bootstrap. Use convention-engineering.

First Pass

  1. Classify the task: design, bootstrap, audit, repair, or migrate.
  2. Identify canonical surface:
    • single repo: repo descriptor
    • multi-repo product: epic descriptor
  3. Inspect before editing:
    • descriptor map
    • leaf descriptors
    • Taskfile/check scripts
    • Slock agent dirs if local projection work is requested
    • repo-local skill roots
  4. State the dependency direction: Slock points to repo; repo does not depend on Slock runtime state.
  5. Apply the smallest repo-owned change and name the verification command.

Core Invariants

  • One mapped Slock agent owns exactly one leaf repo.
  • The descriptor owns the map; Slock memory and symlinks are generated projections.
  • MEMORY.md is pointer-only. Durable rules live in the repo.
  • Default work model is per_leaf_tracked: every leaf owns its own .work/.
  • Cross-repo refs are qualified: <prefix>#IN-NNNN or <prefix>#W-NNNN.
  • Handoffs create target repo inbox/work state; Slock only notifies.
  • Repo-local skills live under the leaf runtime roots actually used by agents.
  • Runtime tokens, session logs, raw channel exports, and copied Slock memory are never canonical repo data.

Workflow By Mode

Design

Use references/protocol.md to choose surfaces, ownership boundaries, work model, and runtime skill discovery. Produce a concrete target tree and registry shape.

Bootstrap

Use references/bootstrap.md to add descriptor fields, leaf domain claims, Taskfile commands, target-owned install/check scripts, and generated Slock projection rules.

Audit

Use references/audit-migration.md to compare descriptor, leaf repos, Slock agent dirs, memory files, notes, and skill roots. Lead with drift and risks.

Repair

Repair repo-owned state first: descriptors, checks, skills, docs, and work handoff records. Regenerate Slock MEMORY.md and symlinks only from the descriptor.

Migrate

Move durable rules from Slock notes/memory into the owning repo surface. Leave historical logs local unless the user explicitly asks to archive them in a repo-owned record.

Approval Points

Ask before:

  • deleting or rewriting historical Slock notes/
  • replacing a non-symlink path inside a Slock agent dir
  • introducing two-way sync between Slock and repo
  • changing agent ownership, repo names, handles, or work prefixes
  • committing local Slock runtime material

Verification

Use references/verification.md. A good change names and runs the relevant gate, usually:

task verify
task slock:check

If the task changes this skill's behavior, update MAINTENANCE.md in the same change. For prior failures, add a regression case entry there; do not create a separate eval/ folder.

References

FileUse For
references/protocol.mdOwnership model, work model, Slock boundary, skill discovery
references/bootstrap.mdDescriptor, Taskfile, generated projection, and script scaffolding
references/audit-migration.mdSlock-dir scan, stale notes, unmapped agents, migration targets
references/verification.mdMechanical gates, success checks, regression evidence
MAINTENANCE.mdSelf-evolution charter and regression cases

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.