Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/threadctx-dev/threadctx-mcp/team-memorynpx skills add threadctx-dev/threadctx-mcp --skill team-memorygit clone --depth 1 https://github.com/threadctx-dev/threadctx-mcpWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00071 | $0.00615 |
| Opus 5 | $0.00036 | $0.00308 |
| Sonnet 5 | $0.00014 | $0.00123 |
| Haiku 4.5 | $0.00007 | $0.00061 |
Grade A, and why
team-memory scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Team memory (threadctx)
threadctx is the team's shared memory. Anything you store can be recalled later by you, by a teammate, or by a teammate's coding agent — across Claude Code, Cursor, and any MCP client. Treat it as writing to a future colleague.
When to query
Call memory_query BEFORE starting work, not after hitting a wall:
- Touching a service, module, or config that looks like it has history
- Fixing a bug that feels like someone may have hit it before
- Making a choice between approaches (a past decision may already settle it)
- Doing anything with deploys, credentials, environments, or CI
Query with a plain-language description of what you're about to do, not keywords: "adding a new billing webhook handler" beats "webhook".
When to write
Call memory_write AFTER:
- Resolving a non-obvious bug (especially one whose symptom pointed the wrong way)
- Making an architectural or tooling decision, including what was rejected and why
- Discovering a gotcha: a flaky step, a misleading error, an undocumented dependency
- Learning something about the environment that isn't in the repo (account quirks, dashboard settings, rate limits)
Don't write things the repo already records (code structure, obvious history) — write the part that ISN'T visible from the code.
How to write a memory a teammate can use
A good memory answers three questions without the reader having your context:
- What happened / what was decided — concrete symptom or decision, with file paths, commands, error text where relevant.
- Why — the root cause or the reasoning, including dead ends ruled out.
- What to do about it — the action a future reader should take or avoid.
Bad: "Fixed the cache bug in the API." Good: "GET /api/v1/memory/list served stale data after PATCH because Next.js Data Cache caches the Neon HTTP driver's fetch() calls even with dynamic='force-dynamic'. Fix: also set fetchCache='force-no-store' on every route using an HTTP DB driver. Verify with a mutate-then-read round-trip against production — the bug only appears on the second read."
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 59 lines · 71 tokens per session scan A 66946a1b61fa
team-memory is a skill published in the GitHub repository threadctx-dev/threadctx-mcp (1 stars, last pushed 1mo ago), licensed MIT. It adds 71 tokens to every session and 615 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…