sentoagent
Advanced tools
+1
-1
| { | ||
| "name": "sentoagent", | ||
| "version": "2.2.1", | ||
| "version": "2.2.2", | ||
| "description": "Agents sent to fight your battles. Self-improving AI agents powered by Claude Code.", | ||
@@ -5,0 +5,0 @@ "author": "Gabriel Gil", |
@@ -40,2 +40,17 @@ export function renderStartAgent(config) { | ||
| while true; do | ||
| # ── Kill any stray channel bridge from a previous session BEFORE launching ── | ||
| # When claude is KILLED (watchdog, OOM, crash) rather than exiting cleanly, its channel | ||
| # bridge child (bun run .../discord|telegram|slack|imessage/...) can be ORPHANED and keep | ||
| # its gateway connection alive. The next loop then spawns a SECOND bridge on the SAME bot | ||
| # token. Discord allows one gateway session per token, so the two connections kick each | ||
| # other and re-IDENTIFY in a tight loop, burning through Discord's hard cap of 1000 | ||
| # IDENTIFYs per token per day in minutes — at which point Discord force-RESETS the token | ||
| # and the agent goes dark until a human pastes a new one. The busiest / most-restarted | ||
| # agent hits this first (it happened repeatedly to one family-assistant agent). | ||
| # | ||
| # -u "$(whoami)" scopes the kill to THIS agent's own bridges, never a sibling's. On a | ||
| # clean first start there is nothing to kill; the sleep lets the socket close before the | ||
| # new bridge dials in, so there is never a two-bridge overlap on one token. | ||
| pkill -u "$(whoami)" -f 'discord/|telegram/|slack/|imessage/' 2>/dev/null && sleep 2 | ||
| claude --dangerously-skip-permissions --model ${config.model || "sonnet"} --channels ${channelPlugin} | ||
@@ -42,0 +57,0 @@ sleep 15 |
AI-detected potential code anomaly
Supply chain riskAI has identified unusual behaviors that may pose a security risk.
Found 7 instances
Long strings
Supply chain riskContains long string literals, which may be a sign of obfuscated or packed code.
URL strings
Supply chain riskPackage contains fragments of external URLs or IP addresses, which the package may be accessing at runtime.
AI-detected potential code anomaly
Supply chain riskAI has identified unusual behaviors that may pose a security risk.
Found 6 instances
Long strings
Supply chain riskContains long string literals, which may be a sign of obfuscated or packed code.
URL strings
Supply chain riskPackage contains fragments of external URLs or IP addresses, which the package may be accessing at runtime.
232919
0.5%4683
0.3%62
1.64%