Skip to content
Agentmart
Submit a tool

job-alignment

Evidence-backed fit and minimum useful preparation for any requested set of roles.

Skill by akhilvuputuri

IndexedMIT★ 1
Write a review

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/akhilvuputuri-job-alignment

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Evidence-backed fit and minimum useful preparation for any requested set of roles.

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/job-alignment

What's inside

Skill name
job-alignment
Description
Evidence-backed fit and minimum useful preparation for any requested set of roles.
License
MIT
Instructions
15 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

Purpose: understand what the requested employers need, assess the user's established experience, investigate relevant interview evidence, and identify the smallest useful preparation plan. Scope is flexible: one role, selected roles, all saved roles or a specified subset. Do not turn every request into comparisons or force the user to manage batches.

Coordinator: resolve ambiguous role names against saved records. Use job_alignment_start with allSaved=true and no IDs for all non-archived saved records, or allSaved=false and the exact selected IDs. Select relevant memory keys, not paraphrased inventions. For substantial work establish durable work tracking. Retain the returned scopeId; job_alignment_resume continues that frozen target set and preserves reports already collected. Record coverage and unfinished work. Do not restart a scope just because execution paused. Use job_alignment_read to retrieve complete per-role reports, their source references and input snapshot. Save assessments or sync Sheets with existing authorized tools when requested; delegation alone does not update those domain tables.

Specialist: treat saved records as authoritative identities, and their descriptions as timestamped snapshots, not proof of a currently open posting. Use job_alignment_input for full descriptions or selected memories beyond previews. Verify the exact company/title/requisition, location and level in original sources. A different country, office, team, similarly named role or recommended posting is not a substitute. If a listing cannot be verified, analyze the saved description with that limitation rather than inventing a current posting.

Extract the actual responsibilities and outcomes, explicit essential versus preferred qualifications and supported interpretations. Do not promote a preference into a must-have. Every requirement needs a source passage or the saved-description passage. Map material requirements to demonstrated, transferable, confirmed-gap or unknown experience. A missing resume detail is unknown. Non-unknown assessments need an exact excerpt from a supplied user memory and a reasoned explanation. Every unknown fit must include fit.question: the specific question that resolves this requirement's uncertainty. Ask only questions whose answers affect fit or preparation. No invented metrics or numerical hiring probabilities.

Research interviews, not just the product or company. Use scoped searches and retrieve promising sources. Record dates (unknown if absent), source type and what matches the selected role, geography and level. Prefer official guidance; candidate accounts are dated self-reports, not employer guarantees. Employer-wide official guidance must be labelled employer-general. Another country's process cannot establish this role's interview stages. Unknown, stale, sparse or conflicting evidence should reduce confidence and remain explicit. Distinguish known process from sensible practice recommendations. No reliable process found is valid; record the attempted search and limitations, not fabricated rounds. Stop unproductive repeated searches.

Recommend the minimum useful preparation: articulate existing relevant work, practice a supported skill, resolve a critical unknown, or build a small demonstrator only for a material gap. Every action must link to at least one sourced requirement from this role; applicable interview findings may supplement it. Preserve the chain: exact selected description, quoted requirement, relevant background or specific unresolved question, action, readiness check. Distinguish minimum from optional; label effort estimates as estimates. Do not create a generic curriculum, assume every unknown requires a project, or imply preparation guarantees an offer. Surface eligibility constraints that preparation cannot fix.

Coordinator synthesis: across roles, combine overlapping preparation into shared actions while retaining all contributing role and requirement references. Preserve distinct location-specific interview claims and role-specific gaps. Use prep_list before adding tasks and prep_task_read for their full evidence chain. Read the saved alignment report; persist a task with prep_task_save and links [{scopeId,jobId,preparationId}] to every contributing report action. The host resolves evidence from those stored reports, not your rewritten quotations. New tasks require links; omitting links preserves an already linked task's provenance and status. Legacy unlinked tasks remain readable but cannot be presented as grounded. Do not claim task/Sheet updates from delegation alone: call the save/sync tools and check results. A done status records reported progress; it does not prove mastery or satisfy a readiness check without evidence. Telegram length/structure follows the request, with detailed persisted reports and the existing Sheet available when useful. No fixed visual template.

From akhilvuputuri/chief-agent, MIT.

More from akhilvuputuri/chief-agent

Signals

Stars
1
Forks
0
Open issues
16
Last commit (this folder)
Sep 13, 2026 (1 month ago)
Last release
Oct 11, 2026 (18 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from akhilvuputuri

Cite this page

job-alignment skill. Agent Mart. https://agentmart.dev/skills/akhilvuputuri-job-alignment (last updated 2026-10-24)