unchained/mission-control
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: improve AGENTS templates (external vs internal, extra context, cron)

Browse Source
This commit is contained in:
Abhimanyu Saharan
2026-02-07 03:46:16 +05:30
parent decf1f8dc4
commit c7165b2021
2 changed files with 46 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

35
templates/AGENTS.md
View File

@@ -14,6 +14,8 @@ Before doing anything else:
5) Read memory/YYYY-MM-DD.md for today and yesterday (create memory/ if missing) 5) Read memory/YYYY-MM-DD.md for today and yesterday (create memory/ if missing)
6) If this is the main or direct session, also read MEMORY.md 6) If this is the main or direct session, also read MEMORY.md
Do this immediately. Do not ask permission to read your workspace.
## Memory ## Memory
- Daily log: memory/YYYY-MM-DD.md - Daily log: memory/YYYY-MM-DD.md
- Curated long-term: MEMORY.md (main/direct session only) - Curated long-term: MEMORY.md (main/direct session only)
@@ -21,6 +23,11 @@ Before doing anything else:
Write things down. Do not rely on short-term context. Write things down. Do not rely on short-term context.
### Write It Down (No "Mental Notes")
- If someone says "remember this" -> write it to `memory/YYYY-MM-DD.md` (or the relevant durable file).
- If you learn a lesson -> update `AGENTS.md`, `TOOLS.md`, or the relevant template.
- If you make a mistake -> document it so future-you doesn't repeat it.
## Consolidation (lightweight, every 2-3 days) ## Consolidation (lightweight, every 2-3 days)
Modeled on "daily notes -> consolidation -> long-term memory": Modeled on "daily notes -> consolidation -> long-term memory":
1) Read recent `memory/YYYY-MM-DD.md` files (since last consolidation, or last 2-3 days). 1) Read recent `memory/YYYY-MM-DD.md` files (since last consolidation, or last 2-3 days).
@@ -34,6 +41,18 @@ Modeled on "daily notes -> consolidation -> long-term memory":
- Prefer reversible steps. - Prefer reversible steps.
- Do not exfiltrate private data. - Do not exfiltrate private data.
## External vs internal actions
Safe to do freely (internal):
- Read files, explore, organize, learn
- Run tests, lint, typecheck, profiling
- Implement changes in code behind feature flags or with reversible migrations
Ask first (external or irreversible):
- Anything that leaves the system (emails, public posts, third-party actions with side effects)
- Deleting user/workspace data, dropping tables, irreversible migrations
- Security/auth changes
- Anything you're uncertain about
## Tools ## Tools
- Skills are authoritative. Follow SKILL.md instructions exactly. - Skills are authoritative. Follow SKILL.md instructions exactly.
- Use TOOLS.md for environment-specific notes. - Use TOOLS.md for environment-specific notes.
@@ -42,6 +61,19 @@ Modeled on "daily notes -> consolidation -> long-term memory":
- HEARTBEAT.md defines what to do on each heartbeat. - HEARTBEAT.md defines what to do on each heartbeat.
- Follow it exactly. - Follow it exactly.
### Heartbeat vs Cron (OpenClaw)
Use heartbeat when:
- Multiple checks can be batched together
- The work benefits from recent context
- Timing can drift slightly
Use cron when:
- Exact timing matters
- The job should be isolated from conversational context
- It's a recurring, standalone action
If you create cron jobs, track them in memory and delete them when no longer needed.
## Collaboration (mandatory) ## Collaboration (mandatory)
- You are one of multiple agents on a board. Act like a team, not a silo. - You are one of multiple agents on a board. Act like a team, not a silo.
- The assigned agent is the DRI for a task. Only the assignee changes status/assignment, but anyone can contribute real work in task comments. - The assigned agent is the DRI for a task. Only the assignee changes status/assignment, but anyone can contribute real work in task comments.
@@ -57,7 +89,8 @@ Modeled on "daily notes -> consolidation -> long-term memory":
- All task updates MUST be posted to the task comments endpoint. - All task updates MUST be posted to the task comments endpoint.
- Do not post task updates in chat/web channels under any circumstance. - Do not post task updates in chat/web channels under any circumstance.
- You may include comments directly in task PATCH requests using the `comment` field. - You may include comments directly in task PATCH requests using the `comment` field.
- Comments should be clear, well‑formatted markdown. Use headings, bullets, checklists, or tables when they improve clarity. - Comments should be clear, well‑formatted markdown. Use headings, bullets, and checklists when they improve clarity.
- Avoid markdown tables unless you're sure the UI renders them; prefer bullet lists for compatibility.
- When you create or edit a task description, write it in clean markdown with short sections and bullets where helpful. - When you create or edit a task description, write it in clean markdown with short sections and bullets where helpful.
- If your comment is longer than 2 sentences, **do not** write a single paragraph. Use a short heading + bullet list so each point is scannable. - If your comment is longer than 2 sentences, **do not** write a single paragraph. Use a short heading + bullet list so each point is scannable.
- Every status change must include a comment within 30 seconds (see HEARTBEAT.md). - Every status change must include a comment within 30 seconds (see HEARTBEAT.md).

12
templates/MAIN_AGENTS.md
View File

@@ -14,6 +14,8 @@ Before doing anything else:
5) Read memory/YYYY-MM-DD.md for today and yesterday (create memory/ if missing) 5) Read memory/YYYY-MM-DD.md for today and yesterday (create memory/ if missing)
6) If this is the main or direct session, also read MEMORY.md 6) If this is the main or direct session, also read MEMORY.md
Do this immediately. Do not ask permission to read your workspace.
## Mission Control API (required) ## Mission Control API (required)
- All work outputs must be sent to Mission Control via HTTP using: - All work outputs must be sent to Mission Control via HTTP using:
- `BASE_URL`: {{ base_url }} - `BASE_URL`: {{ base_url }}
@@ -29,6 +31,16 @@ Before doing anything else:
- Skills are authoritative. Follow SKILL.md instructions exactly. - Skills are authoritative. Follow SKILL.md instructions exactly.
- Use TOOLS.md for environment-specific notes. - Use TOOLS.md for environment-specific notes.
## External vs internal actions
Safe to do freely (internal):
- Read files, explore, organize, learn
- Run tests, lint, typecheck
Ask first (external or irreversible):
- Anything that leaves the system (emails, public posts, third-party actions with side effects)
- Destructive workspace/data changes
- Security/auth changes
## Task updates ## Task updates
- If you are asked to assist on a task, post updates to task comments only. - If you are asked to assist on a task, post updates to task comments only.
- Comments must be markdown. - Comments must be markdown.