AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified Apache-2.0 Self-run

Vault Health Feedback

skill-r007b34r-opencode-obsidian-knowledge-workflow-vault-health-feedback · by r007b34r

>

No reviews yet
0 installs
12 views
0.0% view→install

Install

$ agentstack add skill-r007b34r-opencode-obsidian-knowledge-workflow-vault-health-feedback

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-r007b34r-opencode-obsidian-knowledge-workflow-vault-health-feedback)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
2mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Vault Health Feedback? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

vault-health-feedback

Constraints

  • NEVER scan whole vault; sample strategically
  • NEVER give more than 3 recommendations
  • Check layers in order: workflow -> cognitive -> structural
  • NEVER equate structural mess with system failure (check workflow first)
  • NEVER repair, move, or restructure without explicit confirmation

Trigger Boundary

Use when: vault feels archival, knowledge not returning, system health diagnosis needed, highest-leverage repair unclear, user says "vault is dead" or "nothing comes back."

Do NOT use for:

  • Inbox routing -> inbox-triage
  • Recent-note relationships -> connection-review
  • Week-level synthesis -> weekly-synthesis
  • Direct restructuring before diagnosis
  • Promoting material -> note-promotion

Boundaries

  • Owns: workflow health diagnosis, cognitive return assessment, structural friction detection
  • Delegates to: inbox-triage (root cause is inbox backlog), opencode-context-maintenance (root cause is context drift)
  • Never absorbs: inbox routing, relationship analysis, synthesis, promotion, direct restructuring

Continuations

| Condition | Next skill | |-----------|-----------| | Root cause is inbox backlog | inbox-triage | | Root cause is context drift | opencode-context-maintenance | | Vault paths inaccessible, possible config issue | opencode-context-maintenance | | Vault is deeply archival, needs structural redesign | Escalate to human |

Companion Skill

All vault operations follow obsidian-mcp. Forbidden: obsidian_patch_note, obsidian_append_to_note. Verify every write through readback.

Procedure

Phase 1: Workflow Layer (always check first)

When assessing health metrics, read references/examples.md.

  1. obsidian_list_notes path:"Inbox/" -> inbox freshness (items, age)
  2. obsidian_search_notes query:"type: synthesis" -> synthesis frequency
  3. Check for recent triage/review/promotion activity
  4. Score: active / stale / dead

Phase 2: Cognitive Layer (only if workflow is active or stale)

  1. Sample 5-10 recent notes for reuse signals (links back, references)
  2. Check if themes recur across time windows
  3. Look for notes that were never revisited after creation
  4. Score: returning value / flat / declining

Phase 3: Structural Layer (only if cognitive layer shows issues)

  1. Check for orphan zones, hollow clusters, tag noise
  2. Identify fragmentation patterns
  3. Score: clean / messy but functional / blocking workflow

Phase 4: Recommend

When formatting diagnosis, read references/templates.md.

1-3 highest-leverage interventions, ranked. Report what NOT to do now.

Health Metrics Thresholds (expand for quantitative reference)

Workflow layer:

  • Inbox items >30 days old = stale signal
  • Synthesis frequency 40% unlinked notes = structural concern
  • Tag entropy >50 tags with 4 with declining cognitive > structural blocking
  • Stale workflow > flat cognitive > messy structural
  • Active workflow with messy structure = healthy (do not intervene)

Gotchas

Gotcha 1: Starting from structure

What happens: Agent counts orphans and messy folders first Why it is wrong: Structural mess may be irrelevant if workflow is healthy Correct approach: Always check workflow -> cognitive -> structural in order

Gotcha 2: Too many recommendations

What happens: Agent gives 5-7 improvement suggestions Why it is wrong: No execution focus; user is overwhelmed Correct approach: Maximum 3 interventions, ranked by leverage

Gotcha 3: Diagnosis becomes cleanup

What happens: Agent starts moving/renaming/restructuring during diagnosis Why it is wrong: Unauthorized modification; diagnosis should inform, not act Correct approach: Report findings, recommend interventions, wait for confirmation

Validators (CI only, not runtime)

  • validators/pre-check.sh: Confirms vault is accessible for sampling
  • validators/post-check.sh: Verifies cognitive -> structural)
  • No repair executed without confirmation

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.