agentleFS
AbilityaiAbilityai/cornelius/.claude/skills/manage-watching-domains/SKILL.md

manage-watching-domains

Skill109 starsChanged 40 days agoMITView on GitHub

Manage domain-watch surveillance configs — seed new domains, review scan status and proposals, activate Watch Log proposals into thinking topics, pause domains, and update PIRs or watch queries

The file

Show its rules and text

We split this skill into the rules it gives an agent

23 rules. About 0 are common, 9 are distinctive, and 14 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 14

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read Brain/05-Meta/Watching/WATCHING-REGISTRY.md.

    line 46
  2. Very specific:

    If proposals are waiting: "[N] proposals in Watch Log — run /manage-watching-domains activate [slug] to seed into the incubation loop."

    line 99
  3. Very specific:

    If domains have last_scan: null: "[N] domains have never been scanned — run /domain-watch to establish baselines."

    line 101
  4. Distinctive:

    Purpose: Add a new domain to the surveillance registry.

    line 107
  5. Distinctive:

    If the domain slug is not provided, ask for it.

    line 113
  6. Very specific:

    Use surveillance for broad domains you want to monitor

    line 117
  7. Very specific:

    Use hypothesis for specific questions where you're waiting for evidence

    line 118
  8. Very specific:

    Check that Brain/05-Meta/Watching/[domain-slug].md does NOT already exist.

    line 129
  9. Distinctive:

    Check the slug is not already in the registry.

    line 131
  10. Distinctive:

    Add a row to the domains table:

    line 178
  11. Very specific:

    Update updated and updated_by in frontmatter.

    line 183
  12. Very specific:

    If the proposal is already marked Activated:, warn and stop.

    line 207
  13. Very specific:

    Check that Brain/05-Meta/Thinking/[suggested-slug].md does NOT already exist.

    line 218
  14. Distinctive:

    Check that the slug is not already in the Thinking Registry.

    line 220
  15. Distinctive:

    [APPROVAL GATE]: Wait for response.

    line 234
  16. Very specific:

    Create Brain/05-Meta/Thinking/[suggested-slug].md using the standard incubation-loop template:

    line 238
  17. Distinctive:

    Add a row to the Active Topics table:

    line 281
  18. Very specific:

    Add paused_reason: "[reason]" to frontmatter

    line 343
  19. Very specific:

    Update updated and updated_by

    line 345
  20. Very specific:

    Update updated and updated_by in frontmatter

    line 351
  21. Very specific:

    Read Brain/05-Meta/Watching/[domain-slug].md.

    line 370
  22. Distinctive:

    Update the changed fields in frontmatter

    line 389
  23. Distinctive:

    After completing this skill's primary task, consider tactical improvements:

    line 413

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Manage Watching Domains

Human-in-the-loop manager for the domain-watch surveillance layer. Handles the full lifecycle: seeding new domains for continuous monitoring, reviewing scan status and signal level, activating Watch Log proposals into thinking topics, pausing domains, and updating intelligence requirements.

Companion skills:

  • .claude/skills/domain-watch/SKILL.md — the autonomous scanning engine (runs daily, perceives and proposes)
  • .claude/skills/manage-thinking-topics/SKILL.md — manages the thinking topics that watching domains feed into

Architecture: domain-watch perceives → manage-watching-domains activates → incubation-loop analyses

State Dependencies

SourceLocationReadWriteDescription
Watching RegistryBrain/05-Meta/Watching/WATCHING-REGISTRY.md✓✓Master list of domains + scan status
Domain ConfigsBrain/05-Meta/Watching/[domain-slug].md✓✓Per-domain PIRs, queries, baseline counts
Watch LogBrain/05-Meta/Watching/WATCH-LOG.md✓✓Proposals awaiting activation
Thinking RegistryBrain/05-Meta/Thinking/THINKING-REGISTRY.md✓✓Destination when activating proposals
Thinking FilesBrain/05-Meta/Thinking/[topic-slug].md✓Created during activation

Modes

Invoke with one of five modes. If no mode is provided, default to review.


Mode: review

Purpose: Full status dashboard — what domains are active, when they last scanned, what signal level they found, and what proposals are sitting in the Watch Log awaiting your decision.

Step 1: Read Registry and Date

date '+%Y-%m-%d'

Read Brain/05-Meta/Watching/WATCHING-REGISTRY.md. Parse all domains.

Step 2: Load Domain Summaries

For each domain, read Brain/05-Meta/Watching/[domain-slug].md. Extract:

  • label - human-readable name
  • mode - surveillance or hypothesis
  • last_scan - date of last scan (null = never scanned)
  • notes_at_scan - baseline note count
  • spawn_threshold - delta needed to trigger proposal
  • priority_intelligence_requirements - list of PIRs (count them)
  • watch_queries - list of search terms (count them)

Step 3: Load Watch Log Proposals

Read Brain/05-Meta/Watching/WATCH-LOG.md.

Extract all proposals that are not yet marked **Activated:**. For each:

  • Domain slug
  • Signal level (HIGH/MEDIUM)
  • Scan date
  • Proposed central question
  • Suggested slug
  • Whether it has been activated or not

Step 4: Present Status Report

## Watching Domains: Status Review [DATE]

### Active Domains (N)

| Domain | Mode | Last Scan | Notes @ Scan | Threshold | PIRs | Status |
|--------|------|-----------|-------------|-----------|------|--------|
| [label] | surveillance/hypothesis | YYYY-MM-DD or never | N | +N | N | active |

### Proposals Awaiting Activation (N)

| Domain | Signal | Scan Date | Proposed Question | Suggested Slug |
|--------|--------|-----------|------------------|----------------|

### Paused Domains (N)

| Domain | Paused Since | Reason |
|--------|-------------|--------|

### Summary
- [N] domains under active surveillance
- [N] proposals waiting for your activation decision
- [N] domains never scanned (next domain-watch run will establish baseline)
- Next domain-watch run: daily at 7am UTC

If proposals are waiting: "[N] proposals in Watch Log — run /manage-watching-domains activate [slug] to seed into the incubation loop."

If domains have last_scan: null: "[N] domains have never been scanned — run /domain-watch to establish baselines."


Mode: seed

Purpose: Add a new domain to the surveillance registry.

Argument: domain slug (e.g., quantum-computing-applications)

Step 1: Gather Requirements

If the domain slug is not provided, ask for it. Then gather:

  1. Label - Human-readable domain name (e.g., "Quantum Computing Applications")
  2. Mode - surveillance (watch for accumulation of notes on a topic) or hypothesis (track whether specific claims become answerable)
    • Use surveillance for broad domains you want to monitor
    • Use hypothesis for specific questions where you're waiting for evidence
  3. Watch queries - 2-4 semantic search terms that capture what to look for in the KB (these run against Local Brain Search)
  4. Priority Intelligence Requirements - 2-4 specific questions you want answered when the domain produces new signal. These should be answerable (not vague). Examples:
    • Bad: "What's happening with quantum computing?"
    • Good: "Is there evidence that quantum advantage has been demonstrated for a practical ML task?"
  5. Spawn threshold - How many new notes must accumulate before triggering a proposal (default: 3; use 2 for fast-moving areas, 5 for slow ones)
  6. Why this domain - 1-2 sentences on why it warrants ongoing surveillance
  7. What would trigger incubation - Concrete examples of the kind of signal worth an incubation topic

Step 2: Validate Slug

Check that Brain/05-Meta/Watching/[domain-slug].md does NOT already exist. If it does, warn and stop.

Check the slug is not already in the registry. If it is, warn and stop.

Step 3: Create Domain Config

date '+%Y-%m-%d'

Create Brain/05-Meta/Watching/[domain-slug].md:

---
created: YYYY-MM-DD
updated: YYYY-MM-DD
created_by: claude-sonnet-4-6
updated_by: claude-sonnet-4-6
agent_version: 01.25
slug: [domain-slug]
label: "[Human-readable label]"
mode: [surveillance|hypothesis]
watch_queries:
  - "[primary search term]"
  - "[secondary search term]"
priority_intelligence_requirements:
  - "PIR 1: [specific answerable question]"
  - "PIR 2: [specific answerable question]"
last_scan: null
notes_at_scan: 0
spawn_threshold: [N]
---

# Watching: [Label]

## Why This Domain
[1-2 sentences on why this domain warrants ongoing surveillance]

## What Would Trigger Incubation
[Concrete examples of the kind of signal that would warrant a thinking topic]

## Related Thinking Topics
None yet.

Step 4: Add to Registry

Edit Brain/05-Meta/Watching/WATCHING-REGISTRY.md:

Add a row to the domains table:

| [domain-slug] | [label] | [mode] | null | active |

Update updated and updated_by in frontmatter.

Step 5: Confirm

Report:

  • Domain config created at Brain/05-Meta/Watching/[domain-slug].md
  • Registry updated - domain is now active
  • First domain-watch run will establish the baseline note count (no proposal on first run)
  • Subsequent runs will detect delta against baseline and score signal

Mode: activate

Purpose: Take a Watch Log proposal and create a thinking topic from it, completing the perception → deliberation handoff.

Argument: suggested-slug from a Watch Log proposal (or domain-slug to pick the latest proposal for that domain)

Step 1: Load the Proposal

Read Brain/05-Meta/Watching/WATCH-LOG.md.

Find the proposal matching the slug argument. If multiple proposals exist for the same domain, show them all and ask which to activate.

If the proposal is already marked **Activated:**, warn and stop.

Extract from the proposal:

  • Proposed central question
  • Suggested slug
  • Initial hypotheses (H1, H2) and starting confidence
  • Known evidence (note links and external sources)
  • Signal basis (why this was flagged)

Step 2: Validate No Duplicate

Check that Brain/05-Meta/Thinking/[suggested-slug].md does NOT already exist.

Check that the slug is not already in the Thinking Registry. If so, warn and offer to use a variant slug.

Step 3: Gather Any Additions

Present the proposed central question and hypotheses. Ask:

"Activating: [suggested-slug]

Central question: [question] H1: [statement] (~X%) H2: [statement] (~Y%)

Any changes before creating? (Press enter to accept as-is, or describe revisions)"

[APPROVAL GATE] Wait for response.

Step 4: Create Thinking File

Create Brain/05-Meta/Thinking/[suggested-slug].md using the standard incubation-loop template:

---
created: YYYY-MM-DD
updated: YYYY-MM-DD
created_by: claude-sonnet-4-6
updated_by: claude-sonnet-4-6
agent_version: 01.25
topic: "[exact central question]"
run_count: 0
last_run: null
status: active
---

# Thinking: [Topic Title]

## Central Question
[The exact question from the proposal, with any revisions]

## Why This Question Matters
[1-2 sentences synthesized from the Watch Log signal basis]

## Initial Hypotheses
| Hypothesis | Initial Confidence | Basis |
|-----------|-------------------|-------|
| H1: [statement] | X% | [from proposal known evidence] |
| H2: [statement] | Y% | [from proposal known evidence] |

## Known Evidence (Pre-Run)
[Note links and external sources from the Watch Log proposal]

## Constraints and Assumptions
[Any scope limits worth noting]

---
*Analytical runs will be appended below by the incubation loop*

Step 5: Add to Thinking Registry

Edit Brain/05-Meta/Thinking/THINKING-REGISTRY.md:

Add a row to the Active Topics table:

| [suggested-slug] | [central question] | active | 0 | null |

Update updated and updated_by in frontmatter.

Step 6: Mark Proposal as Activated

Edit Brain/05-Meta/Watching/WATCH-LOG.md:

Find the proposal entry and append under it:

**Activated:** [[suggested-slug]] — YYYY-MM-DD

Update updated and updated_by in frontmatter.

Step 7: Update Domain Config (Related Thinking Topics)

Edit Brain/05-Meta/Watching/[domain-slug].md:

In the Related Thinking Topics section, add:

- [[suggested-slug]] — activated YYYY-MM-DD

Update updated and updated_by in frontmatter.

Step 8: Confirm

Report:

  • Thinking file created at Brain/05-Meta/Thinking/[suggested-slug].md
  • Added to Thinking Registry as active
  • Watch Log entry marked activated
  • Domain config updated with related topic link
  • First incubation-loop run will apply: ACH Audit (move 0)

Mode: pause

Purpose: Temporarily suspend a domain from scanning (without deleting its config or history).

Argument: domain slug

Step 1: Load Domain

Read Brain/05-Meta/Watching/[domain-slug].md. Show:

  • Label, mode, last scan date, note count at last scan
  • Any related thinking topics

Step 2: Confirm Pause

[APPROVAL GATE] Ask:

"Pausing '[label]' will stop domain-watch from scanning this domain on scheduled runs. The config and Watch Log history are preserved. Reason for pausing? (or type 'cancel')"

Capture the reason.

Step 3: Update Domain Config

Edit Brain/05-Meta/Watching/[domain-slug].md:

  • Add paused_reason: "[reason]" to frontmatter
  • Add paused_on: YYYY-MM-DD to frontmatter
  • Update updated and updated_by

Step 4: Update Watching Registry

Edit Brain/05-Meta/Watching/WATCHING-REGISTRY.md:

  • Change the domain's Status column to paused
  • Update updated and updated_by in frontmatter

Step 5: Confirm

Report:

  • Domain config updated with pause reason and date
  • Registry status set to paused
  • To resume: edit the domain config, remove paused_reason and paused_on, set registry status back to active

Mode: update

Purpose: Modify a domain's PIRs or watch queries to sharpen its signal detection.

Argument: domain slug

Step 1: Load Current Config

Read Brain/05-Meta/Watching/[domain-slug].md. Display current:

  • Watch queries (numbered list)
  • Priority Intelligence Requirements (numbered list)
  • Spawn threshold

Step 2: Gather Changes

[APPROVAL GATE] Ask:

"What do you want to update for '[label]'?

  1. Add/remove/edit a watch query
  2. Add/remove/edit a PIR
  3. Change the spawn threshold (or describe the change directly)"

Apply the requested changes. Do not overwrite other frontmatter fields.

Step 3: Write Updated Config

Edit Brain/05-Meta/Watching/[domain-slug].md:

  • Update the changed fields in frontmatter
  • Set updated: YYYY-MM-DD and updated_by: claude-sonnet-4-6

Step 4: Confirm

Report the specific fields changed and their new values.


Error Handling

ErrorRecovery
Watching Registry missingStop; inform user to create it or run /domain-watch first
Domain config missing for registry entryLog warning; show the slug and offer to remove the orphaned registry entry
Slug already exists (seed mode)Stop; show existing file path
No proposals in Watch Log (activate mode)Report; suggest running /domain-watch to scan domains
Proposal already activated (activate mode)Stop; show the thinking topic it was activated into
Domain already paused (pause mode)Show current pause reason; ask if they want to resume instead

Self-Improvement

After completing this skill's primary task, consider tactical improvements:

  • Review execution: Were there friction points, unclear steps, or inefficiencies?
  • Identify improvements: Could the review output, activation flow, or PIR guidance be sharper?
  • Scope check: Only execution changes — NOT changes to the perceive-flag-propose pipeline design
  • Apply improvement (if identified):
    • Edit this SKILL.md with the specific improvement
    • Keep changes minimal and focused
  • Version control:
    • git add .claude/skills/manage-watching-domains/SKILL.md
    • git commit -m "refactor(manage-watching-domains): <brief improvement>"

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.