Skip to content

donaldfilimon/abi / .agents/skills/sea-learning-controller

sea-learning-controller

Verified

Run or inspect SEA (Sparse Evidence Attention) learning through the real `abi complete --learn` path. The agent REPL has only a session-local `/sea` preference and no `/learn` command.

FileKindSize
SKILL.mdskill md1.2 KB
---
name: sea-learning-controller
description: Run or inspect SEA (Sparse Evidence Attention) learning through the real `abi complete --learn` path. The agent REPL has only a session-local `/sea` preference and no `/learn` command.
---

# SEA Learning Controller

Routes SEA learning to the real CLI and keeps the REPL boundary explicit.

## Usage

```bash
SEA_STORE=$(mktemp -d)
ABI_WDBX_PATH="$SEA_STORE" ./target/debug/abi complete --learn "review evidence"
```

## States

- `abi complete --learn` opens the configured durable store, recalls evidence,
  and updates the adaptive modulator.
- `abi complete` without `--learn` uses the base completion route.
- `/sea on|off|status|toggle` in `abi agent tui` is session-local metadata only;
  it reports `live services=off` and never opens WDBX.

## Implementation

Maps to `crates/abi-cli/src/complete.rs` and `crates/abi-sea/src/learn_loop.rs`.
The durable path persists modulator weights under `modulator:weights`; tests and
smokes must use a scratch store and never the user's live `~/.abi` path.

## Skill Integration

There is no `/learn` REPL command. Use `abi complete --learn` for actual evidence
recall; use `/sea` only to inspect or change the session-local REPL preference.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall