claude

vasana-system

A working proof of concept for catching the same structure in places that have nothing to do with each other - and in your own working process: notice such a regularity, capture it, and test each for a real shared mechanism, not surface resemblance. Today it's a growing, hand-kept list of specific observations that resurface when the shape shows up again, each accumulating a count of the times it has recurred. The self-organizing learner the system is built toward is the genuine direction, not yet the state - an early, already-useful step toward it. Memory for forms, not just facts; two local MCP servers (relational-memory + a usage-weighted knowledge graph, experimental). The term "vasana" is grounded in the plugin's axioms.

2 on GitHubby bogheorghiuv2.7.11MITupdated 3 days ago
/

Plugin vasana-system from marketplace ex-cog-dev in bogheorghiu/ex-cog-dev. Commands read from the repo's manifest — 3 days ago.

This plugin runs code on your machine — its hooks fire automatically on events like tool use and it starts local MCP servers. The exact commands are listed below; read them before installing.

What it installs

Skills (10)

  • break-pattern

    Am I trying harder at the wrong thing? - Detects stuckness loops and breaks them. Triggers on (1) explicit stuckness - "going in circles", "keep trying the same thing", "need fresh perspective", (2) implicit stuckness - blame-shifting ("the framework is broken"), false certainty ("I've tried everyth

  • check-assumptions

    Did I actually verify this, or am I assuming? - Challenges technical decisions against behavioral patterns before proceeding. Be direct. Be challenging. Point out BS. Use when (1) making tool/framework/library recommendations, (2) architecture decisions, (3) comparing options, (4) user announces dec

  • find-similar

    Where else does this pattern appear?" - Cross-domain similarity search for observed patterns. Use when (1) a pattern has been noticed and needs verification across domains, (2) checking if a pattern is genuinely novel or already captured, (3) exploring whether a local observation reflects something

  • iterative-loop-engine

    Am I ACTUALLY done, or did I just stop?" - Iterative workflow engine, all-purpose. Use when (1) task has clear completion criteria, (2) single-pass execution is insufficient, (3) quality requires iteration, (4) "keep going until X" patterns apply. Provides the abstract loop; pair with domain-specifi

  • pattern-library

    What patterns keep showing up across unrelated contexts? - Consult and apply the pattern library: behavioral patterns (vasanas) that persist across domains — analysis, argument, research, design, markets, code. Patterns are neutral observations of how behavior self-organizes, NOT anti-patterns: groo

  • record-pattern

    Records observed patterns with proper structure from current or described interactions. Use when (1) user explicitly asks to record a pattern, (2) user approved creating a pattern after vasana skill suggestion, or (3) user runs /pattern-library add command.

  • self-improving-investigation

    Am I investigating, or just confirming what I already believe? Self-correcting research methodology combining blind worker agents, nested iteration loops, and dialectic synthesis. Use when (1) research requires factual certainty not just plausibility, (2) topic has high bias risk, (3) multiple persp

  • temporal-shaping

    What shape does this process need over time?" - Use when designing any time-based process, workflow, or interaction. The pattern that connects envelope design to cognitive choreography. Use when (1) designing learning progressions, (2) planning project phases, (3) structuring conversations or presen

  • test-pattern

    Tests whether patterns actually work — trigger, emerge, and produce value. Use when (1) user wants to verify a pattern before relying on it, (2) user runs /test-pattern command, (3) a pattern was just recorded and needs verification, or (4) user suspects a pattern isn't working as expected.

  • vasana

    Notices when patterns persist across unrelated contexts and suggests capturing them. Use when (1) something shifted during conversation (stuck→unstuck, vague→clear), (2) the shift came from the INTERACTION not just information, (3) a behavioral pattern recurs across domains, (4) conversation reached

Slash commands (1)

  • /vasana-system:pattern-library

    Browse, add, or update patterns in the library

Subagents (1)

  • self-improvement-agent

    Periodically reviews memories, builds insight graph, suggests improvement plans. Use when (1) starting new session and want growth review, (2) after completing significant work, (3) weekly self-audit, (4) user asks "what have you learned" or "how can we improve".

MCP servers (2)

  • relational-memory

    runs: uvx --from git+https://github.com/bogheorghiu/ex-cog-dev#subdirectory=vasana-system/mcp-servers/relational-memory relational-memory

  • edge-graph

    runs: uvx --from git+https://github.com/bogheorghiu/ex-cog-dev#subdirectory=vasana-system/mcp-servers/edge-graph edge-graph

Hooks (2)

  • SessionStart

    "${CLAUDE_PLUGIN_ROOT}/hooks/vasana.sh"

  • PreToolUse

    "${CLAUDE_PLUGIN_ROOT}/hooks/count-skill-firings.sh"

View source on GitHub

More from bogheorghiu/ex-cog-dev

All 4 plugins →

research-toolkit

claude

Even-handed investigation of loaded, contested topics. Research & investigation skills for thinking-with-AI - investigation protocols (deep-investigation, cui-bono / "follow the money", screening a dev job or employer for hidden military/defense ties), text & source deconstruction, frame rotation, and iterative verification that escalates to a recursive adversarial dialectic (generate a claim's exact opposite and test it). The skepticism is symmetric - turned on the consensus, the contrarian take, the model's own bias, and its own conclusions; it assumes no villain in advance (POSIWID: judge a method by what it does, not what it claims). Ships two deliberately basic MCP servers - thin data-access layers, not analysis engines; the skills do the analysis: financial-mcp (stock price, history, company info and standard technical indicators via yfinance, with a local ticker cache) and transparency-mcp (keyless public data - US Congress bills/members/votes via GovTrack, World Bank development indicators, and US nonprofit Form 990 filings via ProPublica).

22 skills · 2 commands · 4 agents · 2 MCP servers · hooks

2

makers-toolkit

claude

Discipline for prompting and co-authoring with AI - nothing to configure, just skills. intrinsic-prompt-design: give the model the reason behind a rule, not just the rule; the reasons part clearly helps, even paired with plain imperatives, though so far only anecdotally (rigorous testing is still worth doing), and its bolder idea - deliberately leaving space for the model's own agency to fill - is the core experimental proposition. system-pilot: engineering discipline for deterministic systems, adapting the original Universal CLAUDE.md Protocol. skill-activation-testing: a two-tier A/B (cheap proxy + live firing test) for whether a description change actually fires, with the validity limit of each tier named, plus a stdlib scorer over claude plugin eval that adds case-level intervals, paired arm deltas and a pre-registered ship rule.

3 skills · 2 agents

2

security-toolkit

claude

Useful friction against obvious mistakes - a basic draft, not a security product (no sandbox; whatever doesn't match a pattern passes through). Threat-detection and dangerous-action-blocking hooks for Claude Code: prompt-injection detection (tiered, with allowlist; covers MCP tool outputs via the * matcher), block dangerous git operations (push to main, force push, reset --hard, --no-verify, etc.), and Desktop Commander input safeguards (block start_process and set_config_value). Also ships an optional, per-user PR-merge guard - block Claude from running `gh pr merge` so a human always merges - that is OFF by default and toggled with the `/pr-merge-guard` command (on/off/status); the `pr-merge-guard` skill explains it and flips it on request. New: the `windows-wsl-security-verification` skill - a guided "am I compromised?" IOC triage + attack-surface reduction for Windows + WSL2 dev boxes after a supply-chain scare (poisoned npm/PyPI package, trojaned extension), with the detection-reading discriminators that prevent false alarms. Windows/WSL-specific; verifies, does not harden; a clean result raises confidence, never proves. Also ships the `pii-gate` skill and installer - a drop-in PII/secret gate for a git repo (pre-commit + pre-push hooks + a CI workflow) that blocks personal names from reaching a remote and chains gitleaks for secrets. It fires on `git init`, adding a remote, or a first push, because pushed history is permanent and the gate only works if it exists before that first push. Everything fails closed, matching is case-insensitive and whole-word in every layer, the CI denylist secret lives in a GitHub Environment limited to the default branch so no branch run can read it, gitleaks in CI ignores inline allow comments and repo-root config, a matched term is never echoed, and the shipped denylist template is EMPTY by design - you fill in a gitignored copy, nothing personal ever lives in the repo.

3 skills · 1 command · hooks

2

Maintain vasana-system?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex plugins4 plugins