chron-mcp 0.1.2 → 0.1.4

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "chron-mcp",
-  "version": "0.1.2",
-  "mcpName": "io.github.nivaya/chron",
+  "version": "0.1.4",
+  "mcpName": "io.github.SirinivasK/chron",
   "description": "Audit-grade timestamped logs for every AI conversation",
   "repository": {
     "type": "git",

package/skills/chron/SKILL.md

@@ -37,6 +37,8 @@ Call `get_session_history` for the returned `session_id`. If prior messages exis
 
 ## Every Exchange
 
+The rule is simple: **any time you output text and wait for input, log it.** This includes mid-task check-ins, not just clean user→assistant→user flows.
+
 1. Call `log_message(session_id, role="user", content=<exact user message>)` → returns T1
 2. Compose your response body (plain text only — no timestamp header)
 3. Call `log_message(session_id, role="assistant", content=<response body>)` → returns T2
@@ -50,6 +52,10 @@ Call `get_session_history` for the returned `session_id`. If prior messages exis
 
 T1 and T2 are real server timestamps — T2 − T1 is the actual generation time. The DB stores clean body content without the header.
 
+**Mid-task check-ins follow the same rule.** If you pause mid-task and ask the user something ("I've done the analysis, what's next?" / "Which approach do you prefer?"), that pause is an exchange — log the assistant message before outputting it, log the user's reply when it arrives. Every turn that crosses the user boundary gets timestamped, regardless of whether the task is "complete."
+
+**Permission grants** — if Claude pauses for a permission prompt and the user allows/denies, log a user message on resume: `"[permission: <tool> <action> — allowed/denied]"`. This is the only visible record of the interaction since permission clicks are not typed messages.
+
 **Do not use `log_exchange` for live conversations** — it captures both timestamps at the same instant and cannot show a real gap. Use it only for batch imports of historical messages.
 
 ## If Chron MCP is unavailable

package/skills/chron.skill.md

@@ -11,6 +11,8 @@ Call `get_session_history` for the returned `session_id`. If prior messages exis
 
 ## Every Exchange
 
+The rule is simple: **any time you output text and wait for input, log it.** This includes mid-task check-ins, not just clean user→assistant→user flows.
+
 1. Call `log_message(session_id, role="user", content=<exact user message>)` → returns T1
 2. Compose your response body (plain text only — no timestamp header)
 3. Call `log_message(session_id, role="assistant", content=<response body>)` → returns T2
@@ -24,6 +26,10 @@ Call `get_session_history` for the returned `session_id`. If prior messages exis
 
 T1 and T2 are real server timestamps — T2 − T1 is the actual generation time. The DB stores clean body content without the header.
 
+**Mid-task check-ins follow the same rule.** If you pause mid-task and ask the user something ("I've done the analysis, what's next?" / "Which approach do you prefer?"), that pause is an exchange — log the assistant message before outputting it, log the user's reply when it arrives. Every turn that crosses the user boundary gets timestamped, regardless of whether the task is "complete."
+
+**Permission grants** — if Claude pauses for a permission prompt and the user allows/denies, log a user message on resume: `"[permission: <tool> <action> — allowed/denied]"`. This is the only visible record of the interaction since permission clicks are not typed messages.
+
 **Do not use `log_exchange` for live conversations** — it captures both timestamps at the same instant and cannot show a real gap. Use it only for batch imports of historical messages.
 
 ## If Chron MCP is unavailable