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 skills add Alexander-Tyagunov/magician --skill weavegit clone --depth 1 https://github.com/Alexander-Tyagunov/magicianWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/alexander-tyagunov/magician/weave)<a href="https://agentmods.dev/skills/alexander-tyagunov/magician/weave"><img src="https://agentmods.dev/badge/skills/alexander-tyagunov/magician/weave/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/alexander-tyagunov/magician/weave"><img src="https://agentmods.dev/badge/skills/alexander-tyagunov/magician/weave.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00089 | $0.00398 |
| Opus 5 | $0.00044 | $0.00199 |
| Sonnet 5 | $0.00018 | $0.00080 |
| Haiku 4.5 | $0.00009 | $0.00040 |
Grade A, and why
weave 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 12d ago.
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.
What it actually says
$weave — Codex Adapter
Read ../../references/codex-adapter.md, then read ../../../skills/weave/SKILL.md and follow the source skill through that Codex adapter. Keep the source skill's gates and non-negotiables.
Codex equivalents:
- Engine — if agent tools are available, compose the delivery per
../../../skills/weave/references/template.md: sequential TDD per unit →$certify→ parallel multi-lens review → adversarially refute every Critical/High → consolidate. Use generic agents with self-contained role prompts;reviewer,security,simplification, andverificationare responsibilities, notmagician:*profiles. Otherwise execute the same stages locally. - Grounding —
kg check/kg init, thenkg query/kg blastto scope each unit; passfile:linepointers into worker prompts, never whole files. - Write gates — implement + test on a branch; never push, open/merge PRs, or do anything destructive without explicit approval. A unit commit requires authorization and must stage only enumerated task-owned files.
- No context loss — every worker prompt self-contained (Goal/Scope/Inputs/Constraints/Return); workers return distilled summaries.
- Plan gate — show units + structure + rough scale and get approval before running a large pipeline (use Codex's question/approval UI).
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.
- 12d ago First seen · 16 lines · 89 tokens per session scan A bfe07c479821
weave is a skill published in the GitHub repository Alexander-Tyagunov/magician (14 stars, last pushed 11d ago), licensed MIT. It adds 89 tokens to every session and 398 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 2026-08-30.
Other skills, from other repositories
collaborating
Hackathon-native multiplayer collaboration mode -- brain-dump together, claim tasks across machines, flag decisions async. Drives the subcommands of /forge:collaborate.
g-init
The single G-Forge front door — run once after installing the plugin. Detects what's here and routes to /g-onboard (existing codebase) or /g-kickoff (new project) for the brief, scaffolds CLAUDE.md (compact G-rules), g-docs/ROADMAP.md (with the Active Session handoff), g-docs/milestones/, g-docs/todo.md, the…
g-onboard
Onboard G-Forge onto an existing codebase. Reads deeply before asking anything — treats existing CLAUDE.md, rules, agents, and task ledgers as first-class inputs. Only interviews for what it genuinely doesn't know yet. Produces or updates g-docs/projectbrief.md.
g-roadmap
Project-manager-driven milestone planner. Gated phases — feature dump → cluster → sequence → premortem & re-prioritize → approve → write. Narrates reasoning at every step, and runs a premortem + re-prioritization of the whole roadmap whenever a milestone is added or modified. Auto-triggers when no g-docs/ROADMAP.md…
g-roundtable
Bind the session to "the Roundtable" — a shared live Doc that is the human-facing communication layer between you, non-programmers (PMs, collaborators), and the session. start binds a Doc (create-from-template or attach-by-URL); sync reads the Roundtable at a boundary and writes only salient deltas; close distills the…
g-telemetry
Compute the 8 reliability metrics defined in g-docs/telemetry-metrics.md, derive a health profile (stable / cautious / defensive / recovery), and write it to .claude/telemetry-profile for adaptive orchestration in /g-execute and /g-review. Read-only on history; never modifies retros, forecasts, or git state.