mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-10-21 16:27:18 +02:00
## Thinking Path The same 2026-04-30 audit that produced PR #4118 (`Invalid session` regex extension) and the ENOTFOUND classifier (#4931) identified a third stuck-session pattern: **13 failures in 7 days, all on a single agent (Ernest)**, with stderr matching: ``` _ApiError: {"error":{"code":400,"message":"The input token count exceeds the maximum number of tokens allowed 1048576","status":"INVALID_ARGUMENT"}} at ChatCompressionService.compress ``` The root cause is that gemini-cli's `ChatCompressionService` blew the 1M token context limit **during its compression step itself**. Resuming the same session ID will hit the same wall on the next attempt — the session is effectively dead the same way it is when "Invalid session identifier" fires (PR #4118). ## What Changed Extends the `isGeminiUnknownSessionError` regex in `parse.ts` with two phrases: - `exceeds\s+the\s+maximum\s+number\s+of\s+tokens` - `input\s+token\s+count\s+exceeds` Both trigger the **existing** fresh-session retry path in `execute.ts:596` — no new code path. Same extension pattern as PR #4118. ## Verification - `npx vitest run --project @paperclipai/adapter-gemini-local` → 14/14 pass (11 in `parse.test.ts` + 3 existing in `execute.remote.test.ts`) - 2 new tests cover the token-overflow patterns - `pnpm --filter @paperclipai/adapter-gemini-local typecheck` → clean - Audit query against `heartbeat_runs.stderr_excerpt` confirms regex matches all 13 occurrences ## Stacking This PR is stacked on top of #4931 (the ENOTFOUND classifier) which adds the `parse.test.ts` file. If #4931 merges first, this PR's diff is just the regex + 2 tests. If this PR is reviewed first, please merge #4931 first to avoid touching the same test scaffolding twice. ## Risks - **Low.** Single-line regex extension. No new code paths. - The session-reset path is well-trodden (PR #4118 in flight). - If a non-Gemini caller produces a stderr containing "exceeds the maximum number of tokens" by coincidence, they would trigger one unnecessary fresh-session retry. Not plausible in the gemini-cli output context where this stderr is sourced. ## Model Used Claude Opus 4.7 (1M context), Anthropic SDK via Claude Code CLI. ## Checklist - [x] Thinking path traces from audit data to single-line regex change - [x] Model specified - [x] No duplicate of planned core work - [x] Tests pass locally - [x] Tests added (2 new) - [x] N/A — server-side regex - [x] Internal pattern; no docs change - [x] Risks documented - [x] Will address Greptile + reviewer comments before merge - [x] I searched the GitHub PR list for similar PRs and confirmed this is not a duplicate (related: #4118 covers the "Invalid session identifier" regex; this PR extends the same regex with token-overflow phrases) 🤖 Generated with [Claude Code](https://claude.com/claude-code) --------- Co-authored-by: Devin Foley <devin@paperclip.ing> Co-authored-by: Paperclip <noreply@paperclip.ing>
242 lines
7.0 KiB
TypeScript
242 lines
7.0 KiB
TypeScript
import { describe, expect, it } from "vitest";
|
|
import { sessionCodec as claudeSessionCodec } from "@paperclipai/adapter-claude-local/server";
|
|
import { sessionCodec as codexSessionCodec, isCodexUnknownSessionError } from "@paperclipai/adapter-codex-local/server";
|
|
import {
|
|
sessionCodec as cursorSessionCodec,
|
|
isCursorUnknownSessionError,
|
|
} from "@paperclipai/adapter-cursor-local/server";
|
|
import {
|
|
sessionCodec as geminiSessionCodec,
|
|
isGeminiSessionUnrecoverableError,
|
|
} from "@paperclipai/adapter-gemini-local/server";
|
|
import {
|
|
sessionCodec as opencodeSessionCodec,
|
|
isOpenCodeUnknownSessionError,
|
|
} from "@paperclipai/adapter-opencode-local/server";
|
|
import { sessionCodec as acpxSessionCodec } from "@paperclipai/adapter-acpx-local/server";
|
|
|
|
describe("adapter session codecs", () => {
|
|
it("normalizes claude session params with cwd", () => {
|
|
const parsed = claudeSessionCodec.deserialize({
|
|
session_id: "claude-session-1",
|
|
folder: "/tmp/workspace",
|
|
prompt_bundle_key: "bundle-1",
|
|
});
|
|
expect(parsed).toEqual({
|
|
sessionId: "claude-session-1",
|
|
cwd: "/tmp/workspace",
|
|
promptBundleKey: "bundle-1",
|
|
});
|
|
|
|
const serialized = claudeSessionCodec.serialize(parsed);
|
|
expect(serialized).toEqual({
|
|
sessionId: "claude-session-1",
|
|
cwd: "/tmp/workspace",
|
|
promptBundleKey: "bundle-1",
|
|
});
|
|
expect(claudeSessionCodec.getDisplayId?.(serialized ?? null)).toBe("claude-session-1");
|
|
});
|
|
|
|
it("normalizes codex session params with cwd", () => {
|
|
const parsed = codexSessionCodec.deserialize({
|
|
sessionId: "codex-session-1",
|
|
cwd: "/tmp/codex",
|
|
});
|
|
expect(parsed).toEqual({
|
|
sessionId: "codex-session-1",
|
|
cwd: "/tmp/codex",
|
|
});
|
|
|
|
const serialized = codexSessionCodec.serialize(parsed);
|
|
expect(serialized).toEqual({
|
|
sessionId: "codex-session-1",
|
|
cwd: "/tmp/codex",
|
|
});
|
|
expect(codexSessionCodec.getDisplayId?.(serialized ?? null)).toBe("codex-session-1");
|
|
});
|
|
|
|
it("normalizes opencode session params with cwd", () => {
|
|
const parsed = opencodeSessionCodec.deserialize({
|
|
sessionID: "opencode-session-1",
|
|
cwd: "/tmp/opencode",
|
|
});
|
|
expect(parsed).toEqual({
|
|
sessionId: "opencode-session-1",
|
|
cwd: "/tmp/opencode",
|
|
});
|
|
|
|
const serialized = opencodeSessionCodec.serialize(parsed);
|
|
expect(serialized).toEqual({
|
|
sessionId: "opencode-session-1",
|
|
cwd: "/tmp/opencode",
|
|
});
|
|
expect(opencodeSessionCodec.getDisplayId?.(serialized ?? null)).toBe("opencode-session-1");
|
|
});
|
|
|
|
it("normalizes cursor session params with cwd", () => {
|
|
const parsed = cursorSessionCodec.deserialize({
|
|
session_id: "cursor-session-1",
|
|
cwd: "/tmp/cursor",
|
|
});
|
|
expect(parsed).toEqual({
|
|
sessionId: "cursor-session-1",
|
|
cwd: "/tmp/cursor",
|
|
});
|
|
|
|
const serialized = cursorSessionCodec.serialize(parsed);
|
|
expect(serialized).toEqual({
|
|
sessionId: "cursor-session-1",
|
|
cwd: "/tmp/cursor",
|
|
});
|
|
expect(cursorSessionCodec.getDisplayId?.(serialized ?? null)).toBe("cursor-session-1");
|
|
});
|
|
|
|
it("normalizes gemini session params with cwd", () => {
|
|
const parsed = geminiSessionCodec.deserialize({
|
|
session_id: "gemini-session-1",
|
|
cwd: "/tmp/gemini",
|
|
});
|
|
expect(parsed).toEqual({
|
|
sessionId: "gemini-session-1",
|
|
cwd: "/tmp/gemini",
|
|
});
|
|
|
|
const serialized = geminiSessionCodec.serialize(parsed);
|
|
expect(serialized).toEqual({
|
|
sessionId: "gemini-session-1",
|
|
cwd: "/tmp/gemini",
|
|
});
|
|
expect(geminiSessionCodec.getDisplayId?.(serialized ?? null)).toBe("gemini-session-1");
|
|
});
|
|
|
|
it("preserves acpx session params required for compatibility checks", () => {
|
|
const parsed = acpxSessionCodec.deserialize({
|
|
sessionKey: "paperclip:company:agent:task:fingerprint",
|
|
runtimeSessionName: "runtime-session-1",
|
|
acpxRecordId: "record-1",
|
|
acpSessionId: "acp-session-1",
|
|
agentSessionId: "agent-session-1",
|
|
agent: "claude",
|
|
cwd: "/tmp/acpx",
|
|
mode: "persistent",
|
|
stateDir: "/tmp/acpx-state",
|
|
configFingerprint: "fingerprint",
|
|
workspaceId: "workspace-1",
|
|
repoUrl: "https://example.com/repo.git",
|
|
repoRef: "main",
|
|
remoteExecution: {
|
|
environmentId: "environment-1",
|
|
leaseId: "lease-1",
|
|
},
|
|
});
|
|
|
|
expect(parsed).toMatchObject({
|
|
sessionKey: "paperclip:company:agent:task:fingerprint",
|
|
runtimeSessionName: "runtime-session-1",
|
|
acpxRecordId: "record-1",
|
|
acpSessionId: "acp-session-1",
|
|
agentSessionId: "agent-session-1",
|
|
agent: "claude",
|
|
cwd: "/tmp/acpx",
|
|
mode: "persistent",
|
|
stateDir: "/tmp/acpx-state",
|
|
configFingerprint: "fingerprint",
|
|
workspaceId: "workspace-1",
|
|
repoUrl: "https://example.com/repo.git",
|
|
repoRef: "main",
|
|
remoteExecution: {
|
|
environmentId: "environment-1",
|
|
leaseId: "lease-1",
|
|
},
|
|
});
|
|
expect(acpxSessionCodec.serialize(parsed)).toEqual(parsed);
|
|
expect(acpxSessionCodec.getDisplayId?.(parsed)).toBe("runtime-session-1");
|
|
});
|
|
});
|
|
|
|
describe("codex resume recovery detection", () => {
|
|
it("detects unknown session errors from codex output", () => {
|
|
expect(
|
|
isCodexUnknownSessionError(
|
|
'{"type":"error","message":"Unknown session id abc"}',
|
|
"",
|
|
),
|
|
).toBe(true);
|
|
expect(
|
|
isCodexUnknownSessionError(
|
|
"",
|
|
"thread 123 not found",
|
|
),
|
|
).toBe(true);
|
|
expect(
|
|
isCodexUnknownSessionError(
|
|
'{"type":"result","ok":true}',
|
|
"",
|
|
),
|
|
).toBe(false);
|
|
});
|
|
});
|
|
|
|
describe("opencode resume recovery detection", () => {
|
|
it("detects unknown session errors from opencode output", () => {
|
|
expect(
|
|
isOpenCodeUnknownSessionError(
|
|
"",
|
|
"NotFoundError: Resource not found: /Users/test/.local/share/opencode/storage/session/proj/ses_missing.json",
|
|
),
|
|
).toBe(true);
|
|
expect(
|
|
isOpenCodeUnknownSessionError(
|
|
"{\"type\":\"step_finish\",\"part\":{\"reason\":\"stop\"}}",
|
|
"",
|
|
),
|
|
).toBe(false);
|
|
});
|
|
});
|
|
|
|
describe("cursor resume recovery detection", () => {
|
|
it("detects unknown session errors from cursor output", () => {
|
|
expect(
|
|
isCursorUnknownSessionError(
|
|
"",
|
|
"Error: unknown session id abc",
|
|
),
|
|
).toBe(true);
|
|
expect(
|
|
isCursorUnknownSessionError(
|
|
"",
|
|
"chat abc not found",
|
|
),
|
|
).toBe(true);
|
|
expect(
|
|
isCursorUnknownSessionError(
|
|
"{\"type\":\"result\",\"subtype\":\"success\"}",
|
|
"",
|
|
),
|
|
).toBe(false);
|
|
});
|
|
});
|
|
|
|
describe("gemini resume recovery detection", () => {
|
|
it("detects unknown session errors from gemini output", () => {
|
|
expect(
|
|
isGeminiSessionUnrecoverableError(
|
|
"",
|
|
"unknown session id abc",
|
|
),
|
|
).toBe(true);
|
|
expect(
|
|
isGeminiSessionUnrecoverableError(
|
|
"",
|
|
"checkpoint latest not found",
|
|
),
|
|
).toBe(true);
|
|
expect(
|
|
isGeminiSessionUnrecoverableError(
|
|
"{\"type\":\"result\",\"subtype\":\"success\"}",
|
|
"",
|
|
),
|
|
).toBe(false);
|
|
});
|
|
});
|