agentleFS
ProjectAJ14ProjectAJ14/eklavya/skills/progress/SKILL.md

progress

Skill29 starsChanged 22 days agoMITView on GitHub

Show the Eklavya mastery map — which projects you have been learning on, what those sessions actually taught, what you skipped, and what is due for review.

22 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use teach and review.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • GoogleSkill

    teach

    Teach the user a new skill or concept, within this workspace.

    Also in 3 projects that use this

  • SentrySkill

    review

    Fixture review skill.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

2 rules. 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 16 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep it under 20 lines. A wall of statistics is not a progress report.

    line 55
  2. Very specific:

    If projects and recent_concepts are both empty, say so plainly in two lines and suggest building something first — the tool has nothing to …

    line 73

“Common” is measured: kept by at least 6 in 10 of the 16 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

/eklavya:progress

Render the learning map. Read-only: do not quiz, do not teach.

  1. get_learner_profile (no domain — you want everything).
  2. get_concept_graph with include_mastery: true only if you need a domain's shape to explain something the profile already flagged. Usually you do not.
  3. get_gate_status if quiz.enforced is set.

What the report has to answer

In this order, because this is the order a learner asks them:

  1. Which project has this been happening on? projects — repo, answers, passed, last_active. Show the basename, not the full path, unless two basenames collide. Compute accuracy yourself and show it as a percentage. A repo of null anywhere means work recorded before the repo column existed: label that row unattributed, never drop it and never guess.
  2. What have I actually learned? recent_concepts — and lead with the context line, not the slug. mason-lifecycle-hooks means nothing; "moved flutter create into pre_gen so brick files overlay it" is the memory that makes the slug mean something. Slug second, in parentheses.
  3. What did I skip? skipped — declined means they chose to pass, dont_know means they blanked and were taught. Say which. This is the actionable list: each comes back once its review is due, so it is a backlog, not a scolding. Wrong answers come back the same way; right answers never do. If it is empty, say so in one clause; it is good news.
  4. What is due? due_for_review — questions declined, blanked or answered wrong whose review date has come (never a correct answer). The count, then the slugs with their tier_to_ask. Already clamped to the level, so do not re-explain tiers.
  5. What is shaky? weak — each with a one-clause gloss of what the concept actually is. A bare slug list is not a report.
  6. What is mastered? known_total, then the first few slugs from known — that list is ordered by score, strongest first, not by date. If it is 0, name the concepts closest to it instead (one more correct rep, not a first) — a zero with no runway attached reads as failure and it is not.

Finish with one concrete next step: /eklavya:quiz when anything is due, /eklavya:learn <topic> when a domain the developer clearly works in is mostly unseen.

Then one line pointing at eklavya dashboard — the same data as a local page, with the history, the per-session record and every question ever asked on a concept, all of which this report deliberately truncates. Mention it once, at the end; it is the deeper view, not a replacement for answering the question they just asked.

How to render it

Keep it under 20 lines. A wall of statistics is not a progress report.

  • Lead with movement, not inventory. The first line is the level, the project, and what changed. Counts of things that have not happened yet go last or nowhere.
  • Never draw a bar whose biggest segment is unseen. A three-part known/learning/unseen bar on a new learner is a chart of absence: every row looks identical and the one number that moved is invisible. Per domain, one line: general — 21 in progress, 0 mastered, 46 not yet seen. Collapse domains with no activity at all into a single trailing line naming them.
  • Tables only when every column is populated for every row. Two columns of zeros is prose pretending to be data.
  • The seed catalogue is large on purpose. Never present the unseen count as a gap to close — nobody is meant to learn all 87 seeded concepts, and framing it that way makes the honest numbers feel like failure.

When there is nothing yet

If projects and recent_concepts are both empty, say so plainly in two lines and suggest building something first — the tool has nothing to show until there is work to ground it in. Do not render empty sections.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.