joakimp/mempalace-toolkit
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(pi): print and JSON mode skip the feed by default (MEMPALACE_FEED_PRINT=1 to keep it)

Browse Source 
After 2f0edf5 a `pi -p` still waited for one full catch-up feed. lastFeedAt
starts at 0, so the first settled turn feeds, and session_shutdown waits for
that feed. On tor-ms22 this took 45 s (prepare 11 s + mine 31 s) for a 3 s
answer. That feed almost never files the run's own transcript, because a
one-shot run rarely passes the exporter's gate (4 messages, 1000 assistant
chars). What it files is the device's other sessions, and the next interactive
feed re-offers those without duplicates. The source_file is the staged path,
fixed per session UUID. An unchanged file is skipped. For a grown file, its
drawers are purged and then refiled.

Now, when ctx.mode is "print" or "json", neither agent_settled nor
session_shutdown feeds. Without ctx.mode, the bridge falls back to
hasUI === false. A stale ctx keeps the old behaviour. RPC mode is long-lived
and feeds as before. MEMPALACE_FEED_PRINT=1 restores feeding in print and JSON
mode, for a device that only ever runs pi -p.

Measured on tor-ms22 with the workspace copy (-ne -nt --no-session -e):
- pi -p: 2.9 s / 2.8 s
- --mode json: 3.8 s
- -p with MEMPALACE_FEED_PRINT=1: 30.8 s; the probe saw the prepare and race
  timers, so the feed ran.

scripts/test-feed-exit.sh now fires agent_settled and then session_shutdown
with a ctx, and gained cases C (print), D (json), E (print +
MEMPALACE_FEED_PRINT=1) and F (rpc). Against 2f0edf5's source, C and D fail
(mines=1, about 1 s wait). All six scripts/test-*.sh pass.
This commit is contained in:
Joakim Persson
2026-10-13 14:51:37 +02:00
parent 2f0edf51cf
commit f590a3ac5b
3 changed files with 90 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files
extensions/pi/README.md
+22 -8
View File
@@ -84,7 +84,9 @@ The bridge feeds this pi installation's own session transcripts into the
palace by itself — no scheduler, no cron, no manual invocation. It fires on
`session_shutdown` (covers quit, `/new`, `/resume`, `/fork`) and on a
debounced `agent_settled` (covers a long session that later crashes, since a
hard kill runs no shutdown handler at all).
hard kill runs no shutdown handler at all). Interactive (TUI) and RPC sessions
feed. Print and JSON mode (`pi -p`, `--mode json`) do not, unless
`MEMPALACE_FEED_PRINT=1`.
The work is split across two processes, and the reason is a hard constraint,
not a style choice: **the palace is single-writer.** A live pi session
@@ -113,6 +115,7 @@ side of the wiring.
| Var | Default | Effect |
|---|---|---|
| `MEMPALACE_FEED` | `1` | Set `0` to disable automatic feeding entirely. |
| `MEMPALACE_FEED_PRINT` | unset | Set `1` to feed in print and JSON mode too (`pi -p`, `--mode json`). They skip the feed by default, because the next interactive session's feed files the same sessions without duplicates. See [Debugging](#pi--p-prints-its-answer-then-takes-minutes-to-exit). |
| `MEMPALACE_FEED_BIN` | `mempalace-pi-session` | Helper to run. |
| `MEMPALACE_FEED_WING` | `wing_conversations` | Target wing — passed to both the exporter and the `mempalace_mine` call. |
| `MEMPALACE_FEED_DEBOUNCE_MS` | `600000` (10 min) | Minimum gap between mid-session (`agent_settled`) feeds. Bounds crash loss to one window instead of a whole session. |
@@ -586,13 +589,24 @@ Fixed 2026-10-13: the timer is cleared when the race settles.
`session_shutdown` resolving, both when the mine answers and when it times
out. Before the fix the first case held node for the whole deadline.
**What is still expected.** The first settled turn of every session runs a
catch-up feed (`lastFeedAt` starts at 0), and `session_shutdown` waits for it.
A `pi -p` therefore still exits only after one full feed: measured at 45 s on
tor-ms22, of which `--prepare` took 11 s and the mine 31 s. A script that
wants `pi -p` to return at once, and does not need this run's transcript
filed now, sets `MEMPALACE_FEED=0`. The next interactive session's feed picks
up the transcript, if it qualifies.
**Print and JSON mode skip the feed (also 2026-10-13).** With only the timer
fix, a `pi -p` still waited for one full feed. The first settled turn of every
session runs a catch-up feed (`lastFeedAt` starts at 0), and `session_shutdown`
waits for it: 45 s on tor-ms22, of which `--prepare` took 11 s and the mine
31 s, for a 3 s answer. That feed almost never files the run's own transcript,
because a one-shot run rarely passes the exporter's quality gate (4 messages,
1000 assistant characters). What it files is the device's *other* sessions, and
the next interactive feed re-offers those anyway. So when `ctx.mode` is
`print` or `json`, the bridge now feeds neither on `agent_settled` nor on
`session_shutdown`. Measured on tor-ms22: `pi -p` 2.9 s and `--mode json`
3.8 s, against 45 s before; `MEMPALACE_FEED_PRINT=1` gives 30.8 s (a smaller
backlog than the 45 s run). RPC mode is long-lived and feeds as before.
Nothing is filed twice when a later session re-offers a transcript. The
`source_file` is the staged path, which is fixed per pi session UUID. The mine
skips an unchanged file, and for a grown file it purges that file's drawers
before refiling. A device that only ever runs `pi -p` sets
`MEMPALACE_FEED_PRINT=1`, or its transcripts are never fed.
## The `Type.Unsafe` gotcha
extensions/pi/mempalace.ts
+30 -3
View File
@@ -33,6 +33,8 @@
* client also means it automatically targets whichever palace this bridge
* is pointed at (local stdio or a shared remote one).
* - MEMPALACE_FEED=0 disable feeding entirely
* - MEMPALACE_FEED_PRINT=1 also feed in print/json mode (`pi -p`,
* `--mode json`), which skip it by default
* - MEMPALACE_FEED_BIN helper to run (default mempalace-pi-session)
* - MEMPALACE_FEED_WING target wing (default wing_conversations)
* - MEMPALACE_FEED_DEBOUNCE_MS min gap between mid-session feeds (default 600000)
@@ -1174,20 +1176,45 @@ export default async function mempalaceExtension(pi: ExtensionAPI) {
return feedInFlight;
}
// Print and JSON mode (`pi -p`, `--mode json`) do NOT feed unless
// MEMPALACE_FEED_PRINT=1. A one-shot run's own transcript almost never meets
// the exporter's quality gate (MIN_MESSAGES=4, MIN_ASSISTANT_CHARS=1000), so
// what such a run would file is the device's OTHER sessions. That is a
// catch-up, because lastFeedAt starts at 0, and session_shutdown waits for
// it: 45 s on tor-ms22 (prepare 11 s + mine 31 s) for a 3 s answer.
// Skipping loses nothing. The source_file is the staged path, fixed per
// session UUID, and the next interactive feed re-offers every qualifying
// session: an unchanged file is skipped, a grown one is purged and refiled.
// RPC mode is long-lived and feeds as before. A pi whose ctx has no `mode`
// falls back to hasUI, which is false only in print/json mode.
const feedInPrintMode = process.env.MEMPALACE_FEED_PRINT === "1";
const isOneShot = (ctx: unknown): boolean => {
try {
const c = ctx as { mode?: string; hasUI?: boolean } | undefined;
const mode = c?.mode;
if (mode === "print" || mode === "json") return true;
return mode === undefined && c?.hasUI === false;
} catch {
return false; // a stale ctx throws on access: keep the old behaviour
}
};
const feedsThisRun = (ctx: unknown): boolean => feedInPrintMode || !isOneShot(ctx);
// Mid-session feed. A hard container kill runs no handler at all, so this is
// what bounds crash loss to one debounce window instead of a whole session.
// Re-mining a grown transcript purges and refiles that source_file, so
// repeated ticks refresh a session's drawers rather than duplicating them.
pi.on("agent_settled", async () => {
pi.on("agent_settled", async (_event, ctx) => {
if (!feedsThisRun(ctx)) return;
if (Date.now() - lastFeedAt < feedDebounceMs) return;
void feedPalace("tick"); // deliberately not awaited: never stall a turn
});
pi.on("session_shutdown", async () => {
pi.on("session_shutdown", async (_event, ctx) => {
// Feed before stopping the client: we are the palace holder, so nothing
// else can mine while we live. pi awaits this handler, so the mine really
// does complete; feedMineTimeoutMs keeps a wedged palace from hanging exit.
await feedPalace("shutdown");
if (feedsThisRun(ctx)) await feedPalace("shutdown");
client.stop();
});
scripts/test-feed-exit.sh
+38 -6
View File
@@ -24,6 +24,14 @@
# handler resolves. Before the fix it lived until the 20 s deadline.
# B. mine never answers, deadline 2 s: the deadline is still a liveness
# guard; the handler resolves at about 2 s and node exits.
# And the print-mode default (2026-10-13): `pi -p` / `--mode json` do not feed.
# C. ctx.mode "print": neither agent_settled nor session_shutdown mines.
# D. ctx.mode "json": the same.
# E. ctx.mode "print" + MEMPALACE_FEED_PRINT=1: mines, as before.
# F. ctx.mode "rpc" (long-lived): mines, as before.
# Every case fires agent_settled first and then session_shutdown, the order a
# real one-turn run sees. In the feeding cases the shutdown joins the tick's
# in-flight feed, so a run mines exactly once.
set -euo pipefail
REPO_ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
SRC="${1:-$REPO_ROOT/extensions/pi/mempalace.ts}"
@@ -93,10 +101,12 @@ PORT="$(cat "$WORK/port")"
[ -n "$PORT" ] || { echo "FAIL: fake palace did not start" >&2; exit 2; }
# ------------------------------------------------------------------ driver ----
# Loads the extension under a stub pi, fires session_shutdown, prints when the
# Loads the extension under a stub pi, fires agent_settled and then
# session_shutdown with the ctx given in DRIVER_CTX, prints when the shutdown
# handler resolved, then RETURNS. Anything still referenced keeps node alive.
cat >"$WORK/driver.mjs" <<'DRIVER'
const t0 = Date.now();
const ctx = JSON.parse(process.env.DRIVER_CTX || "{}");
const handlers = {};
const pi = new Proxy({}, {
get: (_t, k) => k === "on" ? (ev, fn) => { (handlers[ev] ||= []).push(fn); } : () => {},
@@ -105,7 +115,8 @@ const { default: ext } = await import("./ext.mjs");
await ext(pi);
if (!handlers.session_shutdown) { console.log("NO_SHUTDOWN_HANDLER"); process.exitCode = 3; }
else {
for (const fn of handlers.session_shutdown) await fn({ type: "session_shutdown", reason: "quit" }, {});
for (const fn of handlers.agent_settled || []) await fn({ type: "agent_settled" }, ctx);
for (const fn of handlers.session_shutdown) await fn({ type: "session_shutdown", reason: "quit" }, ctx);
console.log(`HANDLER_RESOLVED_MS=${Date.now() - t0}`);
}
process.on("exit", () => console.log(`EXIT_MS=${Date.now() - t0}`));
@@ -114,13 +125,14 @@ DRIVER
failures=0
check() { if [ "$1" = 0 ]; then echo "ok $2"; else echo "FAIL $2"; failures=$((failures + 1)); fi; }
run_case() { # $1 mine delay ms, $2 feed mine timeout ms -> sets HANDLER EXIT RC
run_case() { # $1 mine delay ms, $2 feed mine timeout ms, $3 ctx JSON, [$4 FEED_PRINT] -> sets HANDLER EXIT RC MINES
curl -fsS "http://127.0.0.1:$PORT/delay/$1" >/dev/null
: >"$WORK/calls.log"
set +e
out="$(cd "$WORK" && env -u MEMPALACE_REMOTE_TOKEN -u MEMPALACE_PI_DEVICE \
MEMPALACE_REMOTE_URL="http://127.0.0.1:$PORT/mcp" MEMPALACE_FEED=1 \
MEMPALACE_FEED_BIN="$WORK/fake-feed" MEMPALACE_FEED_MINE_TIMEOUT_MS="$2" \
DRIVER_CTX="$3" MEMPALACE_FEED_PRINT="${4:-}" \
timeout 60 node --no-warnings driver.mjs 2>"$WORK/driver.err")"
RC=$?
set -e
@@ -130,7 +142,7 @@ run_case() { # $1 mine delay ms, $2 feed mine timeout ms -> sets HANDLER EXIT RC
}
# A. normal case: the mine wins the race.
run_case 1000 20000
run_case 1000 20000 '{"mode":"tui","hasUI":true}'
echo "[A] rc=$RC handler=${HANDLER:-?}ms exit=${EXIT:-?}ms mines=$MINES"
check "$([ "$RC" = 0 ] && echo 0 || echo 1)" "A: driver exits cleanly (rc=$RC)"
check "$([ "$MINES" = 1 ] && echo 0 || echo 1)" "A: shutdown really mined the staged source once (not a vacuous pass)"
@@ -139,12 +151,32 @@ check "$([ -n "$EXIT" ] && [ -n "$HANDLER" ] && [ $((EXIT - HANDLER)) -lt 3000 ]
"A: node exits within 3 s of the handler resolving ($(( ${EXIT:-0} - ${HANDLER:-0} )) ms; the 20 s mine deadline must not hold it)"
# B. wedged palace: the deadline must still end the wait.
run_case 60000 2000
run_case 60000 2000 '{"mode":"tui","hasUI":true}'
echo "[B] rc=$RC handler=${HANDLER:-?}ms exit=${EXIT:-?}ms mines=$MINES"
check "$([ "$RC" = 0 ] && echo 0 || echo 1)" "B: driver exits cleanly (rc=$RC)"
check "$([ -n "$HANDLER" ] && [ "$HANDLER" -ge 1900 ] && [ "$HANDLER" -lt 10000 ] && echo 0 || echo 1)" "B: session_shutdown gives up at the 2 s deadline (${HANDLER:-?} ms)"
check "$(grep -q 'feed (shutdown) failed: .*2000ms' "$WORK/driver.err" && echo 0 || echo 1)" "B: the timeout is reported on stderr"
check "$(grep -Eq 'feed \((tick|shutdown)\) failed: .*2000ms' "$WORK/driver.err" && echo 0 || echo 1)" "B: the timeout is reported on stderr"
check "$([ -n "$EXIT" ] && [ -n "$HANDLER" ] && [ $((EXIT - HANDLER)) -lt 3000 ] && echo 0 || echo 1)" \
"B: node exits within 3 s of the handler resolving ($(( ${EXIT:-0} - ${HANDLER:-0} )) ms; the 60 s server delay must not hold it)"
# C/D. one-shot modes skip the feed: no mine, and node exits at once.
for m in print json; do
run_case 1000 20000 "{\"mode\":\"$m\",\"hasUI\":false}"
echo "[$m] rc=$RC handler=${HANDLER:-?}ms exit=${EXIT:-?}ms mines=$MINES"
check "$([ "$RC" = 0 ] && [ "$MINES" = 0 ] && echo 0 || echo 1)" "$m mode: no mine at agent_settled or session_shutdown (mines=$MINES)"
check "$([ -n "$HANDLER" ] && [ "$HANDLER" -lt 900 ] && [ -n "$EXIT" ] && [ $((EXIT - HANDLER)) -lt 3000 ] && echo 0 || echo 1)" \
"$m mode: shutdown does not wait for a feed (${HANDLER:-?} ms) and node exits promptly"
done
# E. the opt-in restores feeding in print mode.
run_case 1000 20000 '{"mode":"print","hasUI":false}' 1
echo "[print+FEED_PRINT=1] rc=$RC handler=${HANDLER:-?}ms exit=${EXIT:-?}ms mines=$MINES"
check "$([ "$RC" = 0 ] && [ "$MINES" = 1 ] && [ -n "$HANDLER" ] && [ "$HANDLER" -ge 900 ] && echo 0 || echo 1)" \
"print mode + MEMPALACE_FEED_PRINT=1: mines once and shutdown waits for it (mines=$MINES, ${HANDLER:-?} ms)"
# F. RPC mode is long-lived (hasUI true) and keeps feeding.
run_case 1000 20000 '{"mode":"rpc","hasUI":true}'
echo "[rpc] rc=$RC handler=${HANDLER:-?}ms exit=${EXIT:-?}ms mines=$MINES"
check "$([ "$RC" = 0 ] && [ "$MINES" = 1 ] && echo 0 || echo 1)" "rpc mode: still mines (mines=$MINES)"
if [ "$failures" = 0 ]; then echo "PASS: all assertions held"; else echo "FAIL: $failures assertion(s) failed"; exit 1; fi