project-context
Skill0 starsChanged 6 months agoMITView on GitHub
Manage the project rolling context file for team-shared codebase knowledge
We split this skill into the rules it gives an agent
3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.
0
Common
Most of the 24 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 28
If PROJECT-CONTEXT.md does not exist and the subcommand is not add:
- Very specific: line 72
When exceeded, /project-context add warns and suggests running cleanup:
- Distinctive: line 164
All entries are current. Nothing to clean up.
“Common” is measured: kept by at least 6 in 10 of the 24 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.
Files
/project-context [subcommand] [arguments]
Manages the project's rolling context file — team-shared knowledge about the codebase that doesn't fit in ADRs or KB entries. Things like "the legacy auth module is fragile", "CI takes 20 min on the full suite", "prod DB has a 5K connection limit we've hit before."
The file lives at PROJECT-CONTEXT.md in the project root (committed, shared
with the team).
Subcommands
| Invocation | What it does |
/project-context add "<entry>" | Add a new context entry |
/project-context cleanup | Review expired/stale entries |
/project-context (no args) | Display all active entries |
Pre-flight
If PROJECT-CONTEXT.md does not exist and the subcommand is not add:
No PROJECT-CONTEXT.md found. Add your first entry with:
/project-context add "<what you've learned about this codebase>"Stop.
File format — PROJECT-CONTEXT.md
# Project Context
Team-shared knowledge about this codebase. Read by vallorcine agents during
scoping and domain analysis. Entries expire after 90 days by default.
Managed by: `/project-context` — do not edit manually.
---
## Active
### <YYYY-MM-DD> — <one-line summary>
**Expires:** <YYYY-MM-DD>
**Scope:** <"global" | module/area name>
<1-3 sentences of detail.>
---
### <YYYY-MM-DD> — CI takes 20 min on full test suite
**Expires:** 2026-06-16
**Scope:** global
Run focused tests during implementation. Only run full suite in refactor step 2f.
---
## Expired
<!-- Entries moved here by /project-context cleanup. Not read by agents. -->Size management
- Active section cap: 50 entries or ~200 lines (whichever is hit first)
- When exceeded,
/project-context addwarns and suggests running cleanup:⚠ Active context has <n> entries (~<n> lines). Consider running: /project-context cleanup - Expired entries in the
## Expiredsection are not counted toward the cap and are not read by pipeline agents
Scope field
global— always read by agents when they load project context<module/area>— only read when the feature being worked on touches that area (matched against constructs in work-plan.md or the feature description)
project-context add "<entry>" — add a new entry
Display:
───────────────────────────────────────────────
📝 PROJECT CONTEXT · add
───────────────────────────────────────────────Step 1 — Parse and confirm
Parse the entry text. Ask:
── New context entry ──────────────────────────
"<entry>"
Scope: global (or type a module/area name to scope it)
Expires: <date 90 days from now> (or type a date, or: never)
Use AskUserQuestion with options:
- "Save"
- "Adjust"
If "Adjust": ask what to change, re-display, confirm again.
### Step 2 — Write
If `PROJECT-CONTEXT.md` doesn't exist, create it with the file header.
Append the new entry to the `## Active` section:
```markdown
### <YYYY-MM-DD> — <one-line summary derived from entry>
**Expires:** <YYYY-MM-DD>
**Scope:** <scope>
<entry text>
---Display:
✓ Added to PROJECT-CONTEXT.md
Expires: <date>
Scope: <scope>project-context cleanup — review expired and stale entries
Display:
───────────────────────────────────────────────
📝 PROJECT CONTEXT · cleanup
───────────────────────────────────────────────Step 1 — Scan entries
Read PROJECT-CONTEXT.md. Parse all entries in the ## Active section.
For each entry, check:
- Is it past its expiry date? →
expired - Is it within 7 days of expiry? →
expiring-soon - Otherwise →
active
Step 2 — Display summary
── Context entries ────────────────────────────
Active: <n>
Expiring soon: <n>
Expired: <n>If no expired or expiring-soon entries:
All entries are current. Nothing to clean up.Stop.
Step 3 — Process expired entries
For each expired entry:
── Expired: <summary> ─────────────────────────
Added: <date> | Expired: <date> | Scope: <scope>
<entry text>
Type: keep (extend 90d) · archive · delete- keep → update
Expires:to 90 days from today - archive → move entry from
## Activeto## Expiredsection - delete → remove entry entirely
Step 4 — Process expiring-soon entries
For each expiring-soon entry:
── Expiring soon: <summary> ───────────────────
Added: <date> | Expires: <date> (<n> days) | Scope: <scope>
<entry text>
Type: keep (extend 90d) · archive · skipSame actions as expired, plus skip to leave it alone.
Step 5 — Summary
───────────────────────────────────────────────
📝 PROJECT CONTEXT · cleanup complete
Extended: <n> Archived: <n> Deleted: <n> Skipped: <n>
Active entries remaining: <n>
───────────────────────────────────────────────project-context (no args) — display all active entries
Display:
───────────────────────────────────────────────
📝 PROJECT CONTEXT
───────────────────────────────────────────────
<n> active entries (<n> global, <n> scoped)
<date> <scope> <summary> <expires>
────── ──────── ────────────────────────────── ─────────
03-16 global CI takes 20 min on full suite 06-16
03-10 jlsm-table 5K connection limit on prod 06-08
03-01 global Legacy auth module is fragile 05-30
<If any expiring within 7 days:>
⚠ <n> entries expiring soon. Run /project-context cleanup to review.
───────────────────────────────────────────────Agent integration
Pipeline agents read PROJECT-CONTEXT.md at these points:
/feature (scoping) — Step 1 (read project config). Read the Active section
of PROJECT-CONTEXT.md if it exists. Use global entries and any scoped entries
matching the feature description to inform the scoping interview. Do not ask
questions that active context entries already answer.
/feature-domains — Step 2 (survey). Check active context entries for
known constraints that affect domain classification. An entry like "prod DB
has a 5K connection limit" is relevant when the feature involves database work.
/feature-plan — Step 1 (load context). Read active context entries
scoped to the modules being planned. Constraints from context entries should
be reflected in the work plan contracts.
Pre-flight staleness check: At pipeline start, if PROJECT-CONTEXT.md
exists and has expired entries, display:
ℹ <n> expired project context entries. Run /project-context cleanup to review.Informational only — never blocks.
Other versions32 projects use a version of this.
Word for word · 0
None found.
Edited · 1
- gootf/software-agent-sop★ 277% same
Wrote their own version · 31
- shinpr/ai-coding-project-boilerplate★ 236same name
- egorfedorov/claude-context-optimizer★ 127same name
- TenureAI/PhD-Zero★ 52same name
- upex-galaxy/agentic-qa-boilerplate★ 26same name
- sergeeey/Claude-cod-top-2026★ 9same name
- perhapsspy/project-context★ 7same name
- aida-core/aida-marketplace★ 4same name
- I-Onlabs/claude-code-skills★ 4same name
- IgorBabikov/Flowcairn★ 4same name
- OpenxIntelligence/Coding-Agent-Skills★ 4same name
21 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
Used alongsideProjects with it also use code-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.
GoogleSkillcode-review
Adversarial systems, concurrency, memory-model, and architecture code review protocol. Use this skill whenever asked to review code, PRs, diffs, or branches in Filament.
Also in 3 projects that use this
In this repository45 other files it gives its agents.
- decisions/CLAUDE.mdCLAUDE.md
- kb/CLAUDE.mdCLAUDE.md
- spec/CLAUDE.mdCLAUDE.md
- work/CLAUDE.mdCLAUDE.md
- script-devSkill
- architectSkill
- auditSkill
- capabilitiesSkill
- curateSkill
- decisionsSkill
- feature-cleanupSkill
- feature-completeSkill
- feature-coordinateSkill
- feature-domainsSkill
- feature-hardenSkill
- feature-implementSkill
- feature-planSkill
- feature-prSkill
- feature-quickSkill
- feature-refactorSkill
- feature-resumeSkill
- feature-retroSkill
- featureSkill
- feature-testSkill
- kbSkill
- researchSkill
- setup-vallorcineSkill
- spec-authorSkill
- spec-backfillSkill
- spec-initSkill
- spec-resolveSkill
- specSkill
- spec-splitSkill
- spec-verifySkill
- spec-writeSkill
- uninstall-vallorcineSkill
- upgrade-vallorcineSkill
- vallorcine-helpSkill
- work-decomposeSkill
- work-planSkill
- work-resumeSkill
- work-runSkill
- workSkill
- work-startSkill
- work-statusSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
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.

