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 JohannsenLum/canvas-api-mcp --skill canvas-grade-checkgit clone --depth 1 https://github.com/JohannsenLum/canvas-api-mcpWrote 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/johannsenlum/canvas-api-mcp/canvas-grade-check)<a href="https://agentmods.dev/skills/johannsenlum/canvas-api-mcp/canvas-grade-check"><img src="https://agentmods.dev/badge/skills/johannsenlum/canvas-api-mcp/canvas-grade-check/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/johannsenlum/canvas-api-mcp/canvas-grade-check"><img src="https://agentmods.dev/badge/skills/johannsenlum/canvas-api-mcp/canvas-grade-check.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 | $0.00078 | $0.00799 |
| Opus 5 | $0.00039 | $0.00400 |
| Sonnet 5 | $0.00016 | $0.00160 |
| Haiku 4.5 | $0.00008 | $0.00080 |
Grade A, and why
canvas-grade-check 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.
How it starts
The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Canvas Grade Check
Read-only analysis skill. Never modify, submit, or post anything: only compute from what the canvas-api-mcp tools return.
Prerequisites
The canvas-api-mcp MCP server must be installed and connected in the current client.
If tool calls to my_courses, my_grades, or list_assignments are not available, tell
the user to install it first: https://mcp.johannsenlum.com/canvas-lms/install
Procedure
-
Identify the course. Call
my_coursesto find the course in question. If the student named it, match it; if ambiguous, ask or list candidates rather than guessing. -
Get current grades. Call
my_gradesfor the identified course to get the current score/grade as Canvas reports it. -
Get assignment detail. Call
list_assignmentsfor the course to get, for every assignment: points_possible and submission/grading state (graded, submitted-ungraded, not submitted, not yet due). -
Compute remaining unearned points.
- Sum
points_possiblefor assignments that are not yet graded (ungraded-submitted, not-yet-due, not submitted): this is the pool of points still in play. - Sum points already earned from graded work (as reported by the tools).
- Do not estimate a score for anything not yet graded: treat it as unresolved potential, not a guessed number.
- Sum
-
State current standing and what's needed:
- Report the current grade/score exactly as
my_gradesreports it. - Report total possible points, points earned so far, and points remaining (ungraded/outstanding), from step 4.
- If asked "what do I need to get X", compute the math to that answer ONLY if points_possible for all remaining work is known and the weighting scheme is either visible via the API or the student's assumption is used explicitly (see below), otherwise present the arithmetic range that is knowable and stop there.
- Report the current grade/score exactly as
-
State weighting assumptions explicitly. Canvas courses may use weighted assignment groups instead of raw points. If the tools do not expose group weights:
- Say plainly that the grading/weighting scheme is not visible via the API.
- Do not invent or assume a weighting scheme (e.g. do not assume "everything is equally weighted" silently). If a raw-points calculation is shown anyway as a rough approximation, label it clearly as "unweighted approximation, may not match the instructor's actual weighting". Never present it as the real grade.
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 · 68 lines · 78 tokens per session scan A fb5aed5cb181
canvas-grade-check is a skill published in the GitHub repository JohannsenLum/canvas-api-mcp (2 stars, last pushed 28d ago), licensed MIT. It adds 78 tokens to every session and 799 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
canvas-cli
Manage Canvas LMS (https://www.instructure.com/canvas) from the terminal with the canvas CLI — courses, assignments, submissions and grading, modules, pages, quizzes, discussions, announcements, users, enrollments, sections, files, and analytics. Use this whenever the user wants to list or create assignments, grade…
canvas-admin-roster
Admin skill for walking the Canvas account hierarchy: list accounts and sub-accounts, see courses and users under each, look up which canned reports are available, and enroll or remove users from a specific course — one action at a time. Trigger phrases include "admin roster", "list accounts", "sub accounts", "account…
canvas-accessibility-sweep
Educator and learning-designer skill for a pre-launch accessibility and broken-link sweep of a Canvas course. Walks the structural accessibility audit and the link audit over a course's pages, assignments, syllabus, announcements, and (optionally) quizzes, then produces a prioritised remediation list you can work…
canvas-office-hours
Educator skill for running office hours through Canvas Scheduler appointment groups. Create and publish sign-up slots for a course, list existing groups and their time slots, and see which students (or student groups) have reserved — one confirmed action at a time. Trigger phrases include "office hours", "appointment…
canvas-peer-review-tracker
Educator skill for tracking peer-review assignments in Canvas. Lists who has been asked to review whom, who has submitted versus who is still pending, and lets you assign new reviewers or send reminder messages — one student at a time. Trigger phrases include "peer reviews", "who hasn't done their peer review", "peer…
canvas-course-pulse
Educator skill for tracking week-over-week course health trends in Canvas. Surfaces assignment performance trends, login activity, engagement gaps, and struggling students across a longer time horizon than a daily check. Trigger phrases include "course pulse", "course health", "course trends", "week-over-week…