chat-recall@0.7.80
CAUTION· install onlyPublisher claims: "Shared team knowledge, leaked-secret alerts, ranked code findings and tracked tasks, across every AI coding tool you use" Observed in sandbox: no network, touches nothing outside the project, writes no memory.
Coverage
Nothing executable in the bundle ran; the trace covers the install step only.- Tested
- every file read, file written and host contacted by what ran
- Not tested
- use of the skill by an agent
What this means for the label: Caution here covers what ran. The server's tools were not exercised with a real credential, so nothing about their behavior is asserted; the credential summary below says what to do with that.
Before you give it your token
- It will ask for
CHAT_RECALL_SCAN_SECRETSCHAT_RECALL_SECRET_STORECHAT_RECALL_TOKENConfiguration, not credentials: BUILDKITE, CHAT_RECALL_AGY_HOME, CHAT_RECALL_APPEND_BATCH_MAX, CHAT_RECALL_AUTO_UPDATE, CHAT_RECALL_BATCHSCAN_MAX_MB, CHAT_RECALL_CHANNEL, CHAT_RECALL_CLAUDE_HOME, CHAT_RECALL_CLI_VERSION, CHAT_RECALL_CODEX_HOME, CHAT_RECALL_CURSOR_HOME, CHAT_RECALL_CURSOR_IDE_HOME, CHAT_RECALL_DATA_DIR, CHAT_RECALL_DERIVED_MAX_MB, CHAT_RECALL_FULL_BUILD_MAX_MB, CHAT_RECALL_GEMINI_HOME, CHAT_RECALL_HERMES_HOME, CHAT_RECALL_LOG_JSON, CHAT_RECALL_LOG_LEVEL, CHAT_RECALL_MAX_BATCH_BYTES, CHAT_RECALL_MAX_INFLIGHT_MB, CHAT_RECALL_MCP_REGISTRATION, CHAT_RECALL_MESSAGE_HOOK, CHAT_RECALL_MESSAGE_HOOK_DEBUG, CHAT_RECALL_NO_BROWSER, CHAT_RECALL_NO_KIT, CHAT_RECALL_NO_NPM_CHECK, CHAT_RECALL_NO_STAR_PROMPT, CHAT_RECALL_OIDC_ISSUER, CHAT_RECALL_OPENCODE_DB, CHAT_RECALL_OPENCODE_DUMP_MAX_BYTES, CHAT_RECALL_PARENT_SESSION, CHAT_RECALL_PI_HOME, CHAT_RECALL_PI_SESSION_DIR, CHAT_RECALL_REDACT_INDEX, CHAT_RECALL_SCAN_TASKS_PER_WORKER, CHAT_RECALL_SCAN_WORKER_HEAP_MB, CHAT_RECALL_SCAN_WORKERS- Where it talks to
- chat-recall-does-not-exist.invalid (its own)
- What we saw when it ran
- no credential files read · no network
- What we could not test
- use of the skill by an agent
Connect it with a scoped credential, under the guard. It asks for CHAT_RECALL_SCAN_SECRETS, CHAT_RECALL_SECRET_STORE, CHAT_RECALL_TOKEN and talks to chat-recall-does-not-exist.invalid. Give the credential to the guard, not the server: the value never enters the server's environment and only travels to the host you bind. What the tools do with a real credential was not observed.
Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.
What changed since 0.7.26
vet-skills diff chat-recall 0.7.26 0.7.80This version reaches for something 0.7.26 did not: 1 new credential variable (CHAT_RECALL_SECRET_STORE). Compared from the two recorded runs, not from source.
- New credential variables
- CHAT_RECALL_SECRET_STORE
The guard can hold an update like this until someone approves it (runtime.on_change: hold). A change that only shows with a real credential, or on a path the run did not reach, is not in a trace and so not in this comparison.
Is [email protected] safe to install?
KillerSkills ran it in the KillerSkills honeypot sandbox (1 scenario; nothing executable in the bundle ran; the trace covers the install step only) on 2026-11-21 and scored it 68/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Auto-update or dynamic fetch configuration; Auto-update or dynamic fetch configuration; Auto-update or dynamic fetch configuration. Summary: no network, touches nothing outside the project, writes no memory.
Behavior trace
Honeypot run · 0 s · 1 scenarios · 0 deviationsStatic findings
export declare function runAutoUpdate(base: string, authHeaders: Record<string, string>, ownVersion: string, deps?: Partial<UpdateDeps>): Promise<UpdateResult>;Fix: Pin versions and sources; vendor anything fetched at install time.export declare function isAutoUpdateEnabled(_edition: string | undefined, flag: string | undefined): boolean;Fix: Pin versions and sources; vendor anything fetched at install time.void runAutoUpdate(base, authHeaders, ownVersion).catch(() => { });Fix: Pin versions and sources; vendor anything fetched at install time.export function isAutoUpdateEnabled(_edition, flag) {Fix: Pin versions and sources; vendor anything fetched at install time.* newer AND auto-update is enabled, downloads the tarball FROM THE SAME ORIGIN,Fix: Pin versions and sources; vendor anything fetched at install time.export declare function planAutoUpdate(base: string, caps: Caps, ownVersion: string, flag: string | undefined): UpdatePlan;Fix: Pin versions and sources; vendor anything fetched at install time.export function planAutoUpdate(base, caps, ownVersion, flag) {Fix: Pin versions and sources; vendor anything fetched at install time.* auto-update checks) must too — otherwise ONE stalled endpoint wedges the wholeFix: Pin versions and sources; vendor anything fetched at install time.hint: 'No keyring access: install libsecret-tools (Debian/Ubuntu: `sudo apt install libsecret-tools`, 'Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.const plan = planAutoUpdate(base, { cli: data.cli }, ownVersion, process.env.CHAT_RECALL_AUTO_UPDATE);Fix: Pin versions and sources; vendor anything fetched at install time.* The watch daemon calls runAutoUpdate() on a cadence: it asks /api/capabilitiesFix: Pin versions and sources; vendor anything fetched at install time.// dist/auto-update.js → dist/../package.json, i.e. the manifest of theFix: Pin versions and sources; vendor anything fetched at install time.const plan = planAutoUpdate(base, { cli: body.cli }, ownVersion, process.env.CHAT_RECALL_AUTO_UPDATE);Fix: Pin versions and sources; vendor anything fetched at install time.* FLOW per session: fetch every candidate → pick/union the fullest → seed theFix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.+ 'Fedora: `sudo dnf install libsecret`). Headless machines have no keyring at all — 'Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.? 'The server you sync to has not rolled it yet — install it with: npm install -g chat-recall@latest'Fix: Pin the exact version (`[email protected]`, `==1.2.3`) so every install gets the bytes that were reviewed.* FLOW per session: fetch every candidate → pick/union the fullest → seed theFix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent."build": "tsc -b && tsc -p tsconfig.daemon.json && node scripts/bundle.mjs && node scripts/postbuild.mjs",Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.* A daemon keeps the code it loaded. Auto-update installs a new version onFix: Pin versions and sources; vendor anything fetched at install time./** Remove our entries from a profile's hooks.json, where earlier versions put them. */Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.* Auto-update (auto-update.ts) handles the normal case, but it can only help aFix: Pin versions and sources; vendor anything fetched at install time.// `https://user:pass@host`, `ssh://`, `ftp://`, `smtp://` and every otherFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.// One broken source must not kill the sync — but a SYSTEMATICALLY failingFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.void runAutoUpdate(base, authHeaders, ownVersion).catch(() => { });Fix: Pin versions and sources; vendor anything fetched at install time.const plan = planAutoUpdate(base, { cli: body.cli }, ownVersion, process.env.CHAT_RECALL_AUTO_UPDATE);Fix: Pin versions and sources; vendor anything fetched at install time.* so the SAME parser runs on archived/shadow container bytes (viaFix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.void runAutoUpdate(base, authHeaders, ownVersion).catch(() => { });Fix: Pin versions and sources; vendor anything fetched at install time.* foreground (no supervisor) → exiting would just kill the user's process;Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.const childEnv = { ...env, [RELAUNCH_GUARD]: "1" };Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.// dominant real-world leak vector (`export` dumps, .env files, `curl -H`,Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.Cost to the agent
Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.
Known vulnerabilities
No advisories in OSV for chat-recall@0.7.80 as of 37 min ago. This is the package itself, not its dependencies.
What would make this Verified
12 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%- Unclaimed publisher≈ +8 on the totalClaim munhq: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
- Auto-update or dynamic fetch configuration≈ +4 on the totalPin versions and sources; vendor anything fetched at install time.
- Privileged sudo command≈ +4 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
- Hidden zero-width characters detected≈ +4 on the totalKeep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
- URL with embedded credentials≈ +4 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
- Access to sensitive file paths≈ +4 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.