claude-mem
claudecodexPersistent memory for Claude Code: the next session starts where the last one left off.
29 skills · 1 MCP server · hooks
Index a codebase, report its issues and gaps, and set it up — context layer, review policy, verification, hooks — in one pass. Ships the rituals, the indexer and the recall MCP brain. Run /setup-repo in a repo to begin.
Plugin cortex from marketplace cortex in marinvch/Cortex. Commands read from the repo's manifest — 7 days ago.
This plugin runs code on your machine — it starts local MCP servers. The exact commands are listed below; read them before installing.
architecture-reviewFind shallow modules worth deepening, report them as an HTML page, then grill through the one the user picks.
audit-notesVault health in one ritual — find and fix orphans, dead links, stale, duplicate or misplaced files and privacy leaks; `score` gives the read-only four-layer score, `graph` rebuilds the navigator. Triggers — "cortex audit", "fix dead links", "score my vault", "rebuild the graph", "what needs a MOC".
challenge-planInterview the user relentlessly about a plan or decision until every branch is resolved. Triggers — "grill me", "stress-test this", "poke holes in this", or a ritual reaching a decision it must not guess.
change-impactWhat breaks if I change this — the files' dependents, how far out, and which no test covers, from the import graph. --against checks a collision with another session's change set, --size recommends one agent or a team. Triggers — "is this edit safe", "which tests to run", "what does this file touch"
clarify-requestSharpen a vague or short prompt ("fix it", "make it faster") before work starts, or when the prompt hook scores its vagueness 4 or higher. Triggers — "optimize this prompt", "what should I have asked".
connect-mcpSet up this machine for Cortex — install the plugin bundle, and connect the live MCP brain so recall and capture work in every project. Per machine, user scope.
create-teamFor the team lead, once — create the shared private team-brain repo, seed its per-project folders and team config, and push it.
daily-planOpen today's vault note and surface the priorities, what's due and what is still open.
debug-issueDiagnose a hard bug or performance regression with a loop that can go red, ranked against the repo's invariants and untested dependents. Triggers — "diagnose", "debug this", or a test or feature that is broken, failing, flaky or slow.
define-termsSharpen a repo's domain language — challenge fuzzy terms, and write its CONTEXT.md glossary and ADRs. Triggers — "what should we call this", "these two words mean the same thing", "record this decision", "write an ADR".
design-specDesign before code for a risky or non-trivial change — brainstorm, then a design spec and a plan, grounded in the repo's Cortex context. Writes no code. Triggers — "spec this", "design before building", "write a design doc".
end-of-day-summaryEnd-of-day consolidation — a dated digest of what changed and what was learned, written into the repo's committed .cortex/memory/. Triggers — "dream", "wrap up the day", "what did we learn today".
handover-notesWrite the state of this session to a handoff file in the OS temp directory, for the next agent to pick up.
join-teamJoin the shared team-brain from a product repo — clone it locally and drop a connector into the repo. Triggers — "connect this repo to the team brain", "join the team brain".
merge-workGet finished work onto main without stranding any — judge it against the repo's docs, one pull request at a time, merge in a safe order, clean up. Triggers — "open a PR", "which PR first", "merge and continue", "clean up the old branches", "качи го", "комитни и push".
migrate-legacy-engineMove a repo off the retired `.ai-os/` engine onto plain-files Cortex without losing its memory — harvest first, delete second.
personal-setupOne-time vault setup — interview the user, fill the context/ files, seed home.md and populate connections.md.
quick-noteDrop a thought, link or idea into the vault inbox in one step, unorganized.
register-projectsSync the vault with the repos on this machine — which have a codebase brain, which are registered in projects/, and register the rest as metadata.
repo-mapRender the repo as one offline HTML page — state overview, import graph, context tree, files, areas and gaps. Triggers — "show me the graph", "visualise this repo", "open the map", "cortex view".
resolve-conflictsResolve a merge conflict or a stuck rebase hunk by hunk, by each side's intent, and finish the merge. Never aborts. Triggers — "resolve these conflicts", "the rebase is stuck", "CONFLICT (content)".
resume-workPick up work in flight — what is committed, uncommitted and diverged, the remaining work, and where the repo is in the Cortex setup sequence, before touching anything. Triggers — "continue where we left off", "what's left", "what's next", "what do I run next", "I installed Cortex, now what", "продъл
retrospectiveBiweekly vault ritual — Notice, Decide, Build, to find one piece of leverage and ship it.
review-changeReview a change against the repo's own AGENTS.md, CONTEXT.md and ADRs — does it break a documented rule, and did it make a doc wrong? Triggers — "review this diff", "check before I commit", "do the docs still match the code".
set-profileShow or set which world this install serves — home, work or lab — which decides what the firewall refuses. Triggers — "this is my work laptop", "which profile am I on", or a refused capture the user disputes.
setup-repoSet up a new or legacy repo — the whole Cortex loop in one pass (root brief, verification, PR review, hooks, evals, agent team) or a report only. Asks once, edits no source. Triggers — "install cortex", "set this repo up", "onboard this project", "just index this", "show me the findings", or a repo
story-mapA plain status of the team's epics and stories, traced to the code — what each epic touches, what waits on an open ask, where the risk is. Never estimates. Triggers — "status of the epic", "which stories are at risk", "what is blocked", "map the stories to the code".
suggest-skillsPropose and write skills that fit this codebase's detected stack — a webhook skill for Stripe, a migration skill for Prisma. Triggers — "add skills for this project", "what skills should this repo have".
summarize-filesAdd summaries, roles and tags on top of the deterministic index, so recall says what code means. Costs tokens, so say so before starting. Triggers — "enrich the index", "what does each file do".
sync-websiteBring Cortex's own public site in line after a release — refresh site-facts.json, redraft the changed pages, open a PR once approved. Triggers — a site-drift issue, "sync the site", "the site is out of date".
trim-agent-docsAudit and slim agent context files (AGENTS.md, CLAUDE.md, rules, skill bodies) in a repo or ~/.claude, and judge a skill collection by use — never reached, redundant. Triggers — "optimize the context", "our AGENTS.md is bloated", "clean up my global CLAUDE.md", "review my skills", "are any of these
update-pluginMake the Cortex this session runs match the repo — check the repo, marketplace and cache copies, update the ones behind, verify the version moved. Triggers — "my change isn't showing up", "update the plugin", "why is it running the old version".
weekly-reviewWeekly vault maintenance — empty the inbox, update projects, restamp current-focus, archive what is stale.
whats-changedSummarize what changed on a project while the user was away, from vault notes and git history. Triggers — "catch me up", "what did I miss", "what changed while I was away".
write-agent-docsThe discipline for writing documents an agent reads — skills, AGENTS.md, CONTEXT.md, any doc behind a pointer — whether a rule belongs in a skill or a section, and wiring a new Cortex skill. Triggers — "this brief is bloated", "why doesn't the agent pick this up", "make a skill for X", "add a ritual
write-area-briefWrite a scoped AGENTS.md for a critical, high-churn or invariant-heavy directory, and wire it into the root routing table. Triggers — "this area needs its own context", "split AGENTS.md", "give billing its own brief".
write-context-docsWrite the context layer into a repo — root AGENTS.md, shims, CONTEXT.md, docs/adr/ and .cortex/. The apply step after a findings report. Triggers — "add the context layer", "scaffold cortex here", "write the AGENTS.md".
cortex-auditorRead-only vault auditor for the Cortex second-brain. Dispatched by the /audit-notes ritual to scan the WHOLE vault in an isolated context and return one ranked findings report covering structure (orphans, dead links, stale/duplicate/misplaced files, integrity/privacy leaks, skill-wiring drift) AND a
cortex-role-reviewerReviews a change from ONE named expert angle — security, performance, accessibility, data integrity, operability or developer experience — grounded in the target repo's own index rather than a generic checklist. Dispatched by /review-change with a role and a diff; returns ranked findings with file:l
cortexruns: node ${CLAUDE_PLUGIN_ROOT}/mcp/server.js
Persistent memory for Claude Code: the next session starts where the last one left off.
29 skills · 1 MCP server · hooks
Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications
5 skills · hooks
Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your coding agent's context.
1 MCP server
Perform AI code review on Git diffs — supports workspace changes, branch ranges, and single commits with concurrent per-file analysis, codebase search, and deep context-aware review.
2 commands
Context-Driven Development plugin that transforms Claude Code into a project management tool with structured workflow: Context → Spec & Plan → Implement
3 skills · 6 commands · 1 agent
Long-term semantic memory for Claude Code, powered by OpenViking
5 skills · 1 command · 1 MCP server · hooks
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.