Agent skill

Conducty Vault Graph

by robertbarclayy in robertbarclayy/conducty

Vault hygiene audit — finds orphaned notes, broken wikilinks, stale context, never-consumed designs, missing index entries, and unused improvement experiments.

MITAuto-check passed

Install Conducty Vault Graph

skills CLI
$ npx skills add robertbarclayy/conducty --skill conducty-vault-graph -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install robertbarclayy/conducty conducty-vault-graph --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/robertbarclayy/conducty.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/conducty-vault-graph .claude/skills/conducty-vault-graph && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
conducty-vault-graph
GitHub stars
176
Token cost
~2.1k tokens
SKILL.md length
724 words
Files
1
Skills in repo
20
Repo updated
First seen
Licence
MIT

At a glance

Vault hygiene audit — finds orphaned notes, broken wikilinks, stale context, never-consumed designs, missing index entries, and unused improvement experiments.

  • Works in 5 steps: Run All Audits → Severity Triage → Write the Audit Report → …
  • The user says vault audit
  • SKILL.md covers When to Use, Audits, Workflow and Cadence, plus 1 more section
  • Calls vault

What it does

Conducty Vault Graph is an agent skill from robertbarclayy/conducty. Vault hygiene audit — finds orphaned notes, broken wikilinks, stale context, never-consumed designs, missing index entries, and unused improvement experiments. Writes a Vault Audit YYYY-MM-DD HHmm.md report. Use when the user says "vault audit", "vault health", "find orphans", "stale context", "vault graph", "what's broken in the vault", or weekly as a hygiene cycle.

Its SKILL.md is about 2.1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The repository describes itself as: Stop context-switching. Start batch-planning. Conducty replaces the fragmented cycle of plan → prompt → wait → get distracted → review → fix → repeat with advanced batch planning. The licence is MIT.

When your agent uses it

  • The user says vault audit
  • Whats broken in the vault
  • Weekly as a hygiene cycle

Example prompts

  • “vault audit”
  • “vault health”
  • “find orphans”
  • “/conducty-vault-graph”

Workflow steps

5 steps, taken from the step headings in SKILL.md.

  1. Run All Audits
  2. Severity Triage
  3. Write the Audit Report
  4. Offer Auto-Fixes
  5. Hand Off

What it can do on your machine

Read from SKILL.md and the folder at commit 64aefd5. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Shell commands in SKILL.md call:

    • vault

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Conducty Vault Graph loads about 2.1k tokens when it runs. Until then it costs about 98 tokens; SKILL.md has 724 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~98
When it runs · the whole SKILL.md, loaded when a task matches
~2.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from robertbarclayy/conducty at commit 64aefd5, republished under its MIT licence (© robertbarclayy). 724 words, ~2,117 tokens.

Download SKILL.mdSave it as .claude/skills/conducty-vault-graph/SKILL.md (or your agent's skills folder).
name
conducty-vault-graph
description
Vault hygiene audit — finds orphaned notes, broken wikilinks, stale context, never-consumed designs, missing index entries, and unused improvement experiments. Writes a `Vault Audit YYYY-MM-DD HHmm.md` report. Use when the user says "vault audit", "vault health", "find orphans", "stale context", "vault graph", "what's broken in the vault", or weekly as a hygiene cycle.
aliases
conducty-vault-graph, vault-graph, vault-audit
tags
conducty/skill, conducty/vault, conducty/maintenance

Conducty Vault Graph — Hygiene + Drift Detection

The vault is now load-bearing. Without periodic hygiene, it accumulates orphans (notes nothing links to), dead wikilinks (links to renamed/deleted notes), stale context (project hubs older than the threshold), unused designs (designs no plan ever consumed), and improvement experiments that nobody evaluated.

This skill runs a structured audit and produces a vault-health report.

[!important] Read [[conducty-obsidian]] first Audit relies on the vault's filename and link conventions.

When to Use

  • Weekly hygiene cycle ("vault audit", "vault health")
  • Before a major plan that will rely heavily on context graph navigation
  • When backlinks in Obsidian start showing "no backlinks" on notes that should have them
  • When [[conducty-context]] flags stale hubs and the user wants the full picture

Audits

Audit 1: Orphan Notes

Notes nothing links to. For each per-instance note (Plan, Design, Improvement, Code Review, Ship Report, Context Refresh, module deep notes), check that at least one other note wikilinks it.

bash
VAULT="${CONDUCTY_VAULT:-$HOME/Obsidian/Conducty}"
for f in "$VAULT"/Plans/Plan*.md "$VAULT"/Designs/Design*.md \
         "$VAULT"/Improvements/Improvement*.md \
         "$VAULT"/"Code Reviews"/"Code Review"*.md \
         "$VAULT"/"Ship Reports"/"Ship Report"*.md; do
    [ -f "$f" ] || continue
    base="$(basename "$f" .md)"
    if ! grep -rl --exclude="$base.md" "\\[\\[$base\\]\\]" "$VAULT" > /dev/null; then
        echo "ORPHAN: $base"
    fi
done

Indexes are expected to link orphans. If a per-instance note isn't in its index, that is the bug — fix the index.

Wikilinks pointing at notes that don't exist. Extract every [[Target]] from every vault note, check the target file exists.

bash
# Wikilinks resolve by basename anywhere in the vault, so search recursively.
grep -rho '\[\[[^]]*\]\]' "$VAULT" --include='*.md' | \
  sed 's/\[\[//;s/\]\]//' | sort -u | \
  while read -r target; do
    # Strip alias suffix: [[Note|Display]] → Note
    target="${target%%|*}"
    if ! find "$VAULT" -type f -name "$target.md" -print -quit | grep -q .; then
      echo "BROKEN: [[$target]]"
    fi
  done

Common causes: a note got renamed without updating callers; a wikilink was hand-typed with a typo; a placeholder template wikilink (e.g. [[Plan YYYY-MM-DD HHmm Topic]]) was never replaced.

Audit 3: Stale Context Hubs

For each Context {Project}.md hub, read last_refreshed from frontmatter. Compare against today.

AgeStatus
≤ 14 daysFresh
14–30 daysStale — recommend refresh
> 30 daysAged — recommend refresh before any plan touches this project

If the hub frontmatter has stale_after_days: <N>, use that instead of 14.

Audit 4: Never-Consumed Designs

A design note that no plan note links to. Either the design was abandoned (delete or archive it) or the plan that should have used it never linked back (fix the plan).

bash
for d in "$VAULT"/Designs/Design*.md; do
    base="$(basename "$d" .md)"
    if ! grep -rl --include='Plan*.md' "\\[\\[$base\\]\\]" "$VAULT" > /dev/null; then
        echo "ORPHAN DESIGN: $base"
    fi
done
Audit 5: Improvement Experiments Never Evaluated

Each Improvements/Improvement YYYY-MM-DD HHmm.md proposes experiments. The next improvement note's ## Prior Experiments Evaluated should reference them. Find improvement notes whose experiments aren't evaluated by any later improvement.

This is the killer signal: experiments exist as ritual but don't change behavior. If audits show this every week, the improvement kata is broken — flag it.

Audit 6: Index Coverage

Every per-instance note should appear in its corresponding index. Check:

  • Plans Index lists every Plans/Plan *.md
  • Designs Index lists every Designs/Design *.md
  • Improvements Index lists every Improvements/Improvement *.md
  • Context Index lists every Context/*/Context *.md hub (not slices)

Note appearing in vault but not in index → index drift. Add the wikilink.

Show full SKILL.md (299 more words)Show less
Audit 7: Failure-Pattern Recurrence

Read Accumulators/Failure Patterns.md. Group entries by symptom. Patterns appearing 3+ times across different plans are systemic — flag them prominently. They warrant a template change ([[conducty-plan]] step 5b) or a process change ([[conducty-improve]]), not another individual fix.

Audit 8: Plan-Without-Closure

A plan note exists without an associated Improvement YYYY-MM-DD HHmm.md written after it. Either the plan never finished (still in flight) or [[conducty-review]] / [[conducty-improve]] were skipped. Surface the gap.

Workflow

Step 1: Run All Audits

Run audits 1-8 against the vault. Collect findings.

Step 2: Severity Triage
SeverityAudit findings
CriticalAudit 2 (broken wikilinks), Audit 7 (recurring failure patterns 3+ times)
ImportantAudit 3 (stale hubs > 30d), Audit 5 (uneval'd experiments), Audit 8 (plans without closure)
MinorAudit 1 (orphans), Audit 4 (never-consumed designs), Audit 3 (stale hubs 14-30d), Audit 6 (index drift)
Step 3: Write the Audit Report

Write Vault Audit YYYY-MM-DD HHmm.md:

markdown
---
type: vault-audit
date: YYYY-MM-DD
time: HHmm
tags: [conducty, conducty/vault-audit]
---

# Vault Audit YYYY-MM-DD HHmm

## Summary

- Total notes: {n}
- Orphans: {n}
- Broken wikilinks: {n}
- Stale hubs: {n} (of which aged: {n})
- Never-consumed designs: {n}
- Uneval'd improvement experiments: {n}
- Index drift: {n}
- Recurring failure patterns: {n}
- Plans without closure: {n}

## Critical

### Broken wikilinks
- `[[X]]` referenced from `Plan 2026-04-27 0930.md`, target missing
- ...

### Recurring failure patterns (3+ occurrences)
- "Vague acceptance criteria" — appears in {N} entries across {M} plans
- ...

## Important

### Stale hubs (> 30 days)
- [[Context My App]] — last refreshed 2026-03-10 (53 days)

### Improvement experiments never evaluated
- [[Improvement 2026-04-15 1830]] — proposed 2 experiments; no later kata references them

### Plans without closure
- [[Plan 2026-04-22 1130 Auth]] — no Improvement note follows

## Minor

### Orphans (no backlinks)
- ...

### Designs never consumed
- ...

### Index drift
- `Plans Index` is missing: [[Plan 2026-04-25 1430 Bug Triage]]
- ...

## Recommended Fixes

1. Repair {N} broken wikilinks (Critical)
2. Refresh {N} stale hubs (run `conducty-context` on each)
3. Evaluate {N} prior experiments in next [[conducty-improve]]
4. Update {N} index entries
5. Investigate recurring failure patterns; promote to template change

## Related

- Index: (no index — audits are run on demand)
- Vault contract: [[conducty-obsidian]]
Step 4: Offer Auto-Fixes

For findings the audit can fix mechanically, ask the user before applying:

  • Index drift: prepend missing wikilinks to the right index
  • Orphan plans: ensure the plan is in Plans Index
  • Broken wikilinks caused by a known rename: offer to find-and-replace via Edit + Bash grep -rl

Never auto-delete. Surface, don't prune.

Step 5: Hand Off

If Critical findings exist, recommend running [[conducty-plan]] with a small remediation plan. If only Minor, log the audit and move on.

Cadence

Recommended: weekly. Set a calendar reminder or bake it into the user's Friday wrap-up.

Principles

  • Hygiene over deletion — surface problems, don't auto-remove anything
  • Severity drives action — Critical findings block reliance on the graph; Minor findings are informational
  • Mechanical first — audits run on filenames and grep, not LLM judgment, so they're cheap and deterministic
  • Audit is also a reflection — recurring failure patterns and uneval'd experiments diagnose process drift, not just file drift

© robertbarclayy, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/conducty-vault-graph of robertbarclayy/conducty.

Open the folder on GitHubat commit 64aefd5

Compare with similar skills

Conducty Vault Graph next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Conducty Vault Graph compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Conducty Vault Graph this skillrobertbarclayy/conducty176—~2.1kAutomated safety check: PassMIT
Obsidian Vault Maintaineropenclaw/openclaw392k1 repos~262Automated safety check: PassMIT
Graphagenticnotetaking/arscontexta3.5k—~4.9kAutomated safety check: NotesMIT
Bigquery Graphgoogle/adk-python22k—~4.8kAutomated safety check: PassApache-2.0
Find Releaseflutter/flutter179k—~545Automated safety check: PassBSD-3-Clause
Code Review Graph Buildertirth8205/code-review-graph32k—~295Automated safety check: PassMIT

Similar skills

  • Obsidian Vault Maintainer

    openclaw/openclaw

    Maintain an Obsidian-friendly memory wiki vault with wikilinks, frontmatter, and official Obsidian CLI awareness.

    392k GitHub starsUsed in 1 repo~262 tokens
    Knowledge ManagementAuto-check passed
  • Graph

    agenticnotetaking/arscontexta

    Interactive knowledge graph analysis. An agent skill from agenticnotetaking/arscontexta.

    3.5k GitHub stars~4.9k tokensUpdated 7 mo ago
    Knowledge ManagementAuto-check: notes
  • Bigquery Graph

    google/adk-python

    Official

    Skill for Graph Query Language (GQL) or SQL/PGQ queries against a property graph.

    22k GitHub stars~4.8k tokensUpdated yesterday
    DatabasesAuto-check passed
  • Find Release

    flutter/flutter

    A skill to find the lowest Dart and Flutter release containing a given commit.

    179k GitHub stars~545 tokensUpdated yesterday
    MobileAuto-check passed
  • Code Review Graph Builder

    tirth8205/code-review-graph

    Builds or incrementally updates the code-review knowledge graph for a repository and reports whether the build succeeded, partly or failed.

    32k GitHub stars~295 tokensUpdated 3 days ago
    DevelopmentAuto-check passed
  • Mini Context Graph

    github/awesome-copilot

    Official

    A persistent, compounding knowledge base combining Karpathy's LLM Wiki pattern with a structured knowledge graph.

    40k GitHub starsUsed in 1 repo~2k tokens
    Knowledge ManagementAuto-check passed

More from robertbarclayy/conducty

All 20 skills in this repo
  • Conducty Checkpoint

    robertbarclayy/conducty

    Quality gate between parallelization groups. An agent skill from robertbarclayy/conducty.

    176 GitHub stars~1.5k tokensUpdated 3 mo ago
    Auto-check passed
  • Conducty Code Review

    robertbarclayy/conducty

    Standalone post-implementation code review across a whole branch or PR.

    176 GitHub stars~2k tokensUpdated 3 mo ago
    Auto-check passed
  • Conducty Context

    robertbarclayy/conducty

    Ingest a project directory into the Obsidian vault as a linked context sub-graph — Architecture, Conventions, Invariants, Hotspots, Tests, Glossary, plus per-bounded-context module notes.

    176 GitHub stars~3.7k tokensUpdated 3 mo ago
    Auto-check passed
  • Conducty Debug

    robertbarclayy/conducty

    Leverage-point analysis for failed prompts. An agent skill from robertbarclayy/conducty.

    176 GitHub stars~1.9k tokensUpdated 3 mo ago
    Auto-check passed
  • Conducty Execute

    robertbarclayy/conducty

    Dispatches Claude Code Task subagents to run prompts from a vault plan note.

    176 GitHub stars~2.5k tokensUpdated 3 mo ago
    Auto-check passed
  • Conducty Improve

    robertbarclayy/conducty

    End-of-plan improvement kata. An agent skill from robertbarclayy/conducty.

    176 GitHub stars~1.5k tokensUpdated 3 mo ago
    Auto-check passed

Questions about Conducty Vault Graph

What does Conducty Vault Graph do?

Vault hygiene audit — finds orphaned notes, broken wikilinks, stale context, never-consumed designs, missing index entries, and unused improvement experiments. Conducty Vault Graph is an agent skill from robertbarclayy/conducty. Vault hygiene audit — finds orphaned notes, broken wikilinks, stale context, never-consumed designs, missing index entries, and unused improvement experiments.

When should I use Conducty Vault Graph?

Conducty Vault Graph fits situations like: the user says vault audit; whats broken in the vault; weekly as a hygiene cycle.

How do I install Conducty Vault Graph in Claude Code?

Run `npx skills add robertbarclayy/conducty --skill conducty-vault-graph -a claude-code`. Or copy the skill folder (skills/conducty-vault-graph in robertbarclayy/conducty) into .claude/skills/conducty-vault-graph in your project. Claude Code loads it when a task matches its description.

How do I install Conducty Vault Graph in Codex?

Run `npx skills add robertbarclayy/conducty --skill conducty-vault-graph -a codex`. Or copy the skill folder (skills/conducty-vault-graph in robertbarclayy/conducty) into .agents/skills/conducty-vault-graph in your project. Codex loads it when a task matches its description.

Can I use Conducty Vault Graph in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add robertbarclayy/conducty --skill conducty-vault-graph -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/conducty-vault-graph, .gemini/skills/conducty-vault-graph, .github/skills/conducty-vault-graph and .opencode/skills/conducty-vault-graph in your project.

What does Conducty Vault Graph need to run?

Going by SKILL.md and its folder, Conducty Vault Graph needs the command-line tools its instructions call (vault).

Does Conducty Vault Graph access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Conducty Vault Graph safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Conducty Vault Graph use?

Conducty Vault Graph is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Conducty Vault Graph use?

About 2.1k tokens (SKILL.md is roughly 8.5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Conducty Vault Graph?

Skills that share tags, products or a category with Conducty Vault Graph: Obsidian Vault Maintainer (openclaw/openclaw, 392k stars), Graph (agenticnotetaking/arscontexta, 3.5k stars), Bigquery Graph (google/adk-python, 22k stars) and Find Release (flutter/flutter, 179k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Conducty Vault Graph?

robertbarclayy (a GitHub user) maintains it in robertbarclayy/conducty, which has 176 GitHub stars. The repository holds 20 skills in this directory. The repository was last updated on June 19, 2026.

Source: robertbarclayy/conducty on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.