donaldfilimon/abi / .agents/skills/run-abi
run-abi
VerifiedBuild, launch, and drive the abi stable-Rust project — the `abi` CLI and the `abi-mcp` JSON-RPC server. Use when asked to run, start, build, smoke-test, or screenshot abi, drive its CLI subcommands, exercise the MCP server over stdio, or confirm a change works in the real binaries (not just the test suite).
smoke.sh
View on GitHub#!/usr/bin/env bash
# run-abi smoke harness: build + exercise the real abi CLI and abi-mcp server.
#
# One command to build both binaries, run representative CLI subcommands, and
# drive the MCP server over stdio with real JSON-RPC. Checks exit codes and
# greps output for expected markers. Writes a full transcript to disk.
#
# Usage: .agents/skills/run-abi/smoke.sh
# From any cwd; it resolves the repo root from its own location.
set -uo pipefail
SCRIPT_DIR=$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)
REPO_ROOT=$(cd -- "$SCRIPT_DIR/../../.." && pwd)
cd "$REPO_ROOT"
ABI="$REPO_ROOT/target/debug/abi"
ABI_MCP="$REPO_ROOT/target/release/abi-mcp"
STORE="$REPO_ROOT/target/debug/smoke-memory.jsonl"
TRANSCRIPT="$REPO_ROOT/target/debug/run-abi-smoke.txt"
# Never inherit a caller's live store. Explicit-path WDBX commands below still
# exercise their scratch store; ordinary completion and MCP stay in memory.
export ABI_WDBX_PATH=:memory:
export ABI_WDBX_PERSIST=0
fail=0
pass=0
# target/debug/ may not exist before the first build; ensure the transcript dir is
# writable up front so every log line lands instead of erroring out per-line.
mkdir -p "$(dirname -- "$TRANSCRIPT")"
log() { printf '%s\n' "$*" | tee -a "$TRANSCRIPT"; }
# run <label> <expected-substring|-> -- <command...>
run() {
local label="$1" expect="$2"; shift 2
[ "$1" = "--" ] && shift
log ""
log "### $label"
log "\$ $*"
local out rc
out=$("$@" 2>&1); rc=$?
printf '%s\n' "$out" | tee -a "$TRANSCRIPT"
log "[exit=$rc]"
if [ "$rc" -ne 0 ]; then
log "FAIL: $label exited $rc"; fail=$((fail+1)); return
fi
if [ "$expect" != "-" ] && ! grep -Fq -- "$expect" <<<"$out"; then
log "FAIL: $label missing expected output: $expect"; fail=$((fail+1)); return
fi
pass=$((pass+1))
}
: > "$TRANSCRIPT"
log "=== run-abi smoke @ $(date) ==="
log "repo: $REPO_ROOT"
log "rustc: $(rustup which rustc 2>/dev/null || echo MISSING)"
# --- Build ---------------------------------------------------------------
run "build cli" "-" -- ./tools/cargo.sh build --locked -p abi-cli --target-dir "$REPO_ROOT/target"
# The launcher prefers release: rebuild that exact profile before exercising it.
run "build mcp" "-" -- ./tools/cargo.sh build --locked --release -p abi-mcp --target-dir "$REPO_ROOT/target"
if [ "$fail" -ne 0 ]; then
log "FATAL: build failed; refusing to exercise stale binaries"
exit 1
fi
if [ ! -x "$ABI" ] || [ ! -x "$ABI_MCP" ]; then
log "FATAL: binaries not produced (abi=$ABI abi-mcp=$ABI_MCP)"
exit 1
fi
# --- CLI surface ---------------------------------------------------------
run "cli help" "Usage: abi" -- "$ABI" help
run "cli backends" "Compute Backends:" -- "$ABI" backends
run "cli scheduler" "source=cli-scheduler-status" -- "$ABI" scheduler status
run "cli complete" "model=claude-fable-5" -- "$ABI" complete "smoke: summarize scheduler status"
run "cli plugin list" "Installed Plugins (" -- "$ABI" plugin list
run "cli agent multi" "MULTI-AGENT RESULTS" -- "$ABI" agent multi "smoke multi"
run "cli agent browser" "orchestration=browser-local" -- "$ABI" agent browser "smoke browser"
# --- WDBX store round-trip ----------------------------------------------
rm -f "$STORE" "$STORE".active && rm -rf "$STORE".v2-*
run "wdbx db init" "initialized empty WDBX" -- "$ABI" wdbx db init "$STORE"
run "wdbx block insert" "appended block" -- "$ABI" wdbx block insert "$STORE" abi '{"note":"smoke checkpoint"}'
run "wdbx query" "blocks" -- "$ABI" wdbx query "$STORE"
# --- MCP server over stdio (real JSON-RPC) -------------------------------
log ""
log "### mcp stdio (initialize + tools/list + tools/call)"
MCP_OUT=$(printf '%s\n' \
'{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05","capabilities":{},"clientInfo":{"name":"smoke","version":"0"}}}' \
'{"jsonrpc":"2.0","id":2,"method":"tools/list"}' \
'{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"scheduler_info","arguments":{}}}' \
'{"jsonrpc":"2.0","id":4,"method":"tools/call","params":{"name":"gpu_status","arguments":{}}}' \
'{"jsonrpc":"2.0","id":5,"method":"tools/call","params":{"name":"plugin_list","arguments":{}}}' \
| "$REPO_ROOT/mcp/launcher.sh" stdio 2>>"$TRANSCRIPT")
mcp_rc=$?
printf '%s\n' "$MCP_OUT" | tee -a "$TRANSCRIPT"
log "[exit=$mcp_rc]"
if [ "$mcp_rc" -eq 0 ]; then
pass=$((pass+1)); log "ok: mcp exited 0"
else
fail=$((fail+1)); log "FAIL: mcp exited $mcp_rc"
fi
for marker in '"serverInfo"' '"name":"gpu_status"' 'scheduler running=' 'backend=' 'plugins count=16'; do
if grep -Fq -- "$marker" <<<"$MCP_OUT"; then
pass=$((pass+1)); log "ok: mcp has $marker"
else
fail=$((fail+1)); log "FAIL: mcp missing $marker"
fi
done
# --- Summary -------------------------------------------------------------
log ""
log "=== summary: pass=$pass fail=$fail ==="
log "transcript: $TRANSCRIPT"
[ "$fail" -eq 0 ] && log "SMOKE OK" || log "SMOKE FAILED"
exit "$fail"
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |